Synthesis, characterization and enantioselective free radical reductions of (1R,2S,5R)-menthyldiphenylgermane and its enantiomer

Article abstract of DOI:10.1016/j.tetasy.2004.06.020

a- a- a- (1R,2S,5R)-Menthyldiphenylgermane and its enantiomer have been prepared in a few steps from germanium tetrachloride. The initial step in this sequence, namely the reaction between germanium tetrachloride and menthylmagnesium chloride, produces menthylgermanium trichloride, which is the exclusive product of this Grignard reaction, presumably due to the bulk of the menthyl group. When used at a low temperature (-78°C) and in conjunction with Lewis acids, such as magnesium salts, these chiral germanes are capable of reducing ester functionalized radicals in high enantioselectivity, but in low-moderate yield. For example, (R)-naproxen ethyl ester was obtained in 15% yield and 99% ee by reaction in toluene of 2-bromonaproxen ethyl ester with (1R,2S,5R)-menthyldiphenylgermane in toluene at -78°C in the presence of magnesium bromide. At 80°C, (1R,2S,5R)-menthyldiphenylgermane reacted with primary alkyl radicals with a rate constant of 1.02×10⁶ M^-1 s^-1. Kinetic studies reveal the Arrhenius expression for this reaction to be: log(k/M^-1 s^-1) = (11.1 ± 0.4) - (34.6±3.1)/θ where θ=2.3RT kJ mol^-1.

Full text of DOI:10.1016/j.tetasy.2004.06.020

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 2547–2554
Synthesis, characterization and enantioselective free radical
reductions of (1R,2S,5R)-menthyldiphenylgermane
and its enantiomer
a
Le Zeng, Dainis Dakternieks, Andrew Duthie, V. Tamara Perchyonok and
b
b
a
a,
*
Carl H. Schiesser
a
School of Chemistry, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Victoria, 3010 Australia
b
Centre for Chiral and Molecular Technologies, Deakin University, Geelong, Victoria, 3217 Australia
Received 9 June 2004; accepted 15 June 2004
Abstract—(1R,2S,5R)-Menthyldiphenylgermane and its enantiomer have been prepared in a few steps from germanium tetrachlo-
ride. The initial step in this sequence, namely the reaction between germanium tetrachloride and menthylmagnesium chloride, pro-
duces menthylgermanium trichloride, which is the exclusive product of this Grignard reaction, presumably due to the bulk of the
menthyl group. When used at a low temperature (ꢀ78ꢀC) and in conjunction with Lewis acids, such as magnesium salts, these chiral
germanes are capable of reducing ester functionalized radicals in high enantioselectivity, but in low-moderate yield. For example,
(
(
R)-naproxen ethyl ester was obtained in 15% yield and 99% ee by reaction in toluene of 2-bromonaproxen ethyl ester with
1R,2S,5R)-menthyldiphenylgermane in toluene at ꢀ78ꢀC in the presence of magnesium bromide. At 80ꢀC, (1R,2S,5R)-menthyldi-
6
ꢀ1 ꢀ1
phenylgermane reacted with primary alkyl radicals with a rate constant of 1.02·10 M
s
. Kinetic studies reveal the Arrhenius
.
ꢀ
1
ꢀ1
ꢀ1
expression for this reaction to be: log(k/M
2004 Elsevier Ltd. All rights reserved.
s
)=(11.1± 0.4)ꢀ(34.6± 3.1)/h where h=2.3RTkJmol
ꢁ
1
. Introduction
of providing single enantiomer outcomes for a variety
of transformations of synthetic and commercial signifi-
5
,6
Free-radical chemistry has benefited greatly by the
introduction of chain-carrying reagents such as tributyl-
tin hydride. Indeed, the impact of tin-based reagents in
cance.
An example of this chemistry is shown in
Scheme 1, in which (R)-naproxen ethyl ester 6 is pre-
pared in 99% enantioselectivity (ee) from racemic
bromide 7 by reduction with bis[(1R,2S,5R)-men-
1
this area is hard to overstate; stannanes have allowed the
development of elegant syntheses of novel ring systems
that in turn have led, more recently, to cascade reactions
for the tandem constructions of multi-ring systems of
thyl]phenyltin hydride (Men PhSnH) 2 at ꢀ78ꢀC in
2
6
the presence of magnesium bromide.
MenPh2SnH
1
–3
biological interest.
While this chemistry can now be
1
Me2N
Sn(H)PhMen
applied routinely by the synthetic practitioner, the abil-
ity to control the stereochemical outcomes of free-radi-
cal reactions has provided a greater challenge that has
Men2PhSnH
NMe2
= Men
2
Sn(H)PhMen
3
–6
only recently been adequately addressed.
4
5
Men3SnH
3
Recently, we reported the preparation of chiral, non-
racemic, (1R,2S,5R)-menthyl (Men) containing stan-
nanes 1–5 for use in enantioselective free-radical
reduction chemistry and demonstrated that in con-
junction with Lewis acids, these stannanes are capable
CH3
O
7
,8
Br
OEt Men2PhSnH
MgBr2 / –78˚C
OEt
O
MeO
MeO
7
6
99% ee (R)
*
Corresponding author. Tel.: +61-3-9344-4180; fax: +61-3-9347-
180; e-mail: carlhs@unimeld.edu.au
5
Scheme 1.
0
957-4166/$ - see front matter ꢁ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.06.020

2548
L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
The dominance of trialkylstannanes as the reagents of
choice in free radical chemistry is now being challenged.
ride followed by further synthetic manipulation. It is
possible to react the analogous germane, namely triphen-
ylgermanium chloride, with Grignard reagents. De-
spite this, triphenylgermanium chloride was inert to
menthylmagnesium chloride, even after considerable
experimentation with variables such as solvent, temper-
ature and reaction time performed. It quickly became
apparent that an alternative strategy would be required.
9
10
19
Perceived toxicity and product purification concerns
have lead to the development of ꢀfriendlierꢁ reagents;
1
1–13
tris(trimethylsilyl)silane
1
and tris(trimethyl)silane-
are representative of a new generation of rea-
1,12,14
thiol
gents designed specifically for use in free-radical chain
reactions. Improvements in synthetic flexibility resulting
from the use of germanes and thiols as hydrogen donors
has resulted in extensions of free-radical methodology to
systems in which the primary bond-forming reaction (e.g.
intramolecular addition) lies outside the acceptable range
The literature suggests that Grignard reagents (RMgX)
react with GeCl to provide complex mixtures of the
4
type R GeCl
that require difficult fractional distill-
ation techniques to separate and therefore give low
n
4ꢀn
2
for stannane chemistry. The use of trialkylgermanes, for
2
0,21
example, with typically lower rate constants for the deliv-
ery of hydrogen atom to alkyl radicals can often lead to
increased reaction yields when slow C–C bond-forming
yields of alkylgermanium trichloride.
Despite this,
we chose to react (1R,2S,5R)-menthylmagnesium chlo-
ride directly with germanium tetrachloride. Surprisingly,
this reaction proceeded smoothly to afford (1R,2S,5R)-
menthylgermanium trichloride 10, which was isolated
in 50% yield after simple distillation (Scheme 2). Tri-
chloride 10 proved to be laevorotatory, with
1
5–17
reactions are crucial in the overall synthetic strategy.
With these considerations in mind, Curran et al. recently
described the preparation of chiral germanes (e.g. 8)
based on the chiral bis(naphthalenethiol) unit and
reported enantioselectivities of up to 42% in certain
25
½aꢁ ¼ ꢀ51:3 (c 1, CHCl3).
D
1
8
reductions reaction.
It is curious that this reaction should proceed so well
given that the analogous tin reaction cannot be control-
2
2
led to give a single 1:1 product, and that GeCl usually
4
TMS
S
Ge
S
reacts indiscriminately or poorly with these types of rea-
gents, as described above. We postulate that this is due
to the steric environment of the menthyl substituent,
which leads to the poor reactivity of the Grignard rea-
gent with triphenylgermanium chloride and that this
same steric phenomenon is responsible for the excellent
selectivity observed in the reaction involving GeCl4.
tBu
H
TMS
8
As part of our ongoing development of new reagents
for stereoselective free radical chemistry, we herein
report the development of both enantiomers of ment-
hyldiphenylgermane 9; these chiral germanes are capable
of reducing ester functionalized radicals in high enantio-
selectivity, but in low-moderate yield. Given that a
knowledge of rate constants is crucial to the design of
free-radical reactions of synthetic significance, we have
also determined these important parameters for our chi-
ral germanium reagents. We also report that at 80ꢀC,
Further synthetic manipulation involving treatment of
10 with 3equiv of phenylmagnesium bromide afforded
(1R,2S,5R)-menthyltriphenylgermane 11 in excellent
isolated yield (93%) (Scheme 2). Germane 11 also
25
proved to be laevorotatory {½aꢁ ¼ ꢀ34:8 (c 1, CHCl )}.
D
3
Crystallization from hexane at 4ꢀC afforded germane 11
as a crystalline solid suitable for single-crystal X-ray
analysis, the results of which are displayed in Figure 1.
(
1R,2S,5R)-menthyldiphenylgermane reacts with pri-
mary alkyl radicals with a rate constant of 1.02·
GeCl4
PhMgBr
6
0 M
ꢀ1 ꢀ1
s . Kinetic studies reveal the Arrhenius expres-
1
sion for this reaction to be: log(k/M
ꢀ
1 ꢀ1
s
MgCl
GeCl3
GePh3
)=(11.1±
As ob-
ꢀ
1
0
.4)ꢀ(34.6± 3.1)/h where h=2.3RTkJmol
served previously for related stannanes, menthyl substi-
tution would appear to affect both energy and entropy
terms of this Arrhenius expression.
10
11
Br2
LiAlH4
Ph
Ge H
Ph
Ge Br
2
. Results and discussion
Ph
Ph
2
.1. Synthesis of (1R,2S,5R)-menthyldiphenylgermane 9
9
12
Recently, we developed and published a synthetic meth-
odology for the preparation of (1R,2S,5R)-menthyldi-
phenyltin hydride 1 and its related stannanes. In the
preparation of 1, this methodology involves the reaction
of menthylmagnesium chloride with triphenyltin chlo-
Scheme 2.
7
Inspection of Figure 1 reveals 11 to be quite ordinary,
with regular C–Ge bond lengths and C–Ge–C angles.

L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
2549
Figure 1. Perspective diagram of (1R,2S,5R)-(ꢀ)-menthyltriphenylgermane 11 (left), and its enantiomer ent-11 (right) with key atom numbering.
The most important information provided by the crys-
MenPh2Ge
kc
tallographic information is confirmation that each ster-
eogenic centre in the menthyl substituent is indeed of
the correct configuration; in particular no epimerization,
has occurred during the Grignard chemistry described
above.
Br
kH
9
9
Slow addition of 1equiv of elemental bromine to a solu-
tion of 11 gave (1R,2S,5R)-menthyldiphenylgermanium
bromide 12, which crystallized on standing. Germane
14
13
1
2 was not purified, but further reacted with lithium alu-
minium hydride to provide the required germane,
1R,2S,5R)-menthyldiphenylgermane 9 as a colourless
d[13]
d[14]
kc
rate equation:
=
kH[9]
(
Scheme 3.
oil in 87% yield after distillation (Scheme 2). The struc-
ture of 9 was confirmed by spectroscopic techniques
and microanalytical data and also proved to be laevoro-
tatory {½aꢁD ¼ ꢀ32:2 (c 1, toluene)}.
22
expression in that solvent to be similar to the expressions
determined in other solvents, namely:
As expected, the enantiomer of 9, namely (1S,2R,5S)-
menthyldiphenylgermane ent-9 was prepared in an iden-
tical fashion starting from (1S,2R,5S)-menthyl chloride.
ꢀ
1
log k
c
=s ¼ ð10:13 ꢂ 0:42Þ ꢀ ð27:6 ꢂ 2:6Þ=h
ð1Þ
ꢀ
1
where h is 2.3RTkJmol . We therefore have chosen to
use this expression for the ꢀclock reactionꢁ throughout
this work.
2.2. Determination of kinetic parameters for radical
reactions of 9 with primary alkyl radicals
In order to support our experimental techniques, reac-
tions of chiral germane 9 with 1-bromo-5-hexene were
initially performed at 80ꢀC under ꢀpseudo first-orderꢁ
conditions at three germane concentrations (0.05, 0.1,
0.15M) as described in our previous work. In addition
to this, we also chose to examine the reaction of tributyl-
germane, a reagent that has been studied previously, al-
beit not in tert-butylbenzene. We expected our data for
The absolute rate constants for the delivery of the hydro-
gen atom from 9 to primary alkyl radicals in tert-butyl-
benzene were determined through application of the
2
4
2
3
well-established ꢀ5-hexenyl radical clockꢁ reaction as de-
2
4
scribed by us (Scheme 3). Provided that the ꢀclockꢁ rate
constant (k ) is well established for any given temperature,
then the rate equation (Scheme 3) will provide a value for
the hydrogen transfer rate constant (k ). It should be
c
Bu GeH to correlate with those reported previously, pro-
3
viding a further test for our methodology. Application of
the integrated rate equation (Scheme 3) provided the rate
constant data listed in Table 1. The data presented for
each entry are the average of three individual experi-
ments. The degree of convergence between the data ob-
tained for each system in this manner indicates that the
kinetic model (Scheme 3) is correct and that we are mon-
itoring free-radical processes, as well as supporting our
experimental technique. In addition, the rate constant
H
noted that several published Arrhenius parameters exist
2
5–32
for the ring-closure of the 5-hexenyl radical.
EPR spectroscopy and competitive experiments provide
Kinetic
Arrhenius expressions, with values of 9.5–10.7 for log(A/
s
ꢀ
1
ꢀ1 30,31
,
), and activation energies of 25.5–32.6kJmol
3
2
with the ꢀbestꢁ values being 10.4± 0.3 and 28.7± 1.8. In
our previous work, we calibrated the ꢀhexenyl radical
clockꢁ in tert-butylbenzene and determined the Arrhenius
24

2550
L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
Table 1. Selected rate data for the reaction of primary alkyl radicals with (1R,2S,5R)-menthyldiphenylgermane 9 and tributylgermanane in tert-
butylbenzene at 80ꢀC
ꢀ1 a
)
6
ꢀ1 ꢀ1
Germane
MenPh GeH 9
Temp (ꢀC)
[Germane] (M)
% 13
% 14
k
H
/k
c
(M
k
H
(·10 M
s )
2
80
0.05
0.10
0.15
0.10
95.6
91.5
87.4
97.1
4.4
8.5
0.92
0.93
0.95
0.30
1.01
1.02
1.04
0.33
8
8
0
0
12.6
2.9
Bu
3
GeH
80
a
Average of three experiments.
5
ꢀ1 ꢀ1
s determined for Bu GeH at this
temperature is in excellent agreement with the literature
(
k ) of 3.3·10 M
values of 9.07± 0.24 and 15.45± 1.34 for logA and en-
ergy terms, respectively, for the analogous reaction
involving tributyltin hydride (Bu SnH). It appears
3
H
3
5
ꢀ1 ꢀ1
16
32
value of 3.4·10 M
s
determined in octane.
therefore, that as was observed for reactions involving
the tin-based reagents 1–5, the sterically-demanding
menthyl substituent also exerts an influence on both
the energy and entropy of activation in reactions involv-
ing 9. As was speculated for the tin-based reagents 1–5,
chiral germane 9, with an activation energy of
The remaining kinetic data for reactions at temperatures
other than 80ꢀC were also obtained under ꢀpseudo first-
orderꢁ conditions. Systematic variations in temperature
(
60–120ꢀC) reveal a linear correlation between logkH
and reciprocal temperature for each germane. All kinetic
data are averages of three experiments and errors in
logA and activation energy (E ) are expressed to 95%
ꢀ
1
34.6kJmol for its reaction with the 5-hexenyl radical,
some 15kJmol higher than that that for the analogous
ꢀ
1
a
confidence and account for random, but not systematic
errors. The Arrhenius data obtained in this manner are
summarized in Table 2 together with the available data
reaction involving Bu GeH, must react with a signifi-
3
cantly more disordered transition state than Bu GeH
3
in order to achieve a faster delivery of the hydrogen
atom to the primary alkyl radical than Bu GeH.
3
3
for triphenylgermane.
3
Inspection of Table 2 reveals that the Arrhenius data ob-
tained for the chemistry involving Bu GeH in tert-butyl-
benzene are in excellent agreement with those previously
2.3. Enantioselective reactions of 9 with selected prochiral
radicals
3
1
6
determined in octane, indeed they are identical to
within experimental error, once again supporting our
experimental techniques. The data presented for the
menthyl-substituted system 9 are interesting; it would
appear that 9 reacts with the 5-hexenyl radical with a
significantly greater logA value, as well as a significantly
In our previous work, we described how appropriately
chosen Lewis acid additives can enhance enantioselecti-
vities (ees) during reductions involving chiral stannanes
and radicals, which contain proximate Lewis acid bind-
ing sites, such as ester functionalized systems. It was also
demonstrated that magnesium salts such as MgBr were
2
larger activation energy (E ) than the analogous process
a
involving Bu GeH. It is difficult to make comparisons
3
especially effective in this regard. We have speculated
that the addition of magnesium salts allows for ꢀcoordi-
nated dimers to form, the result being that one ester unit
provides the steric bulk that affords high selectivity dur-
with other germanes in this regard because so little Arr-
henius data exist for these hydrides, and none, to the
best of our knowledge, for systems that bear sterically
6
ing the reduction of the other unitꢁ. Other workers have
used chiral Lewis acids to engender high enantiopurity
1
7
demanding ligands such as menthyl. However, com-
parisons can be made with the available data reported
in free-radical reduction and C–C bond formation
3,4
chemistry, as well as diastereoselective processes.
3
4
for related stannanes. For example, menthyldiphenyl-
tin hydride 1 has been measured to react with the
With this in mind, we set out to determine firstly,
whether or not chiral germanes such as 9 could provide
superior results to their tin counterparts in the absence
of any additive, and secondly, whether or not Lewis acid
additives had the same enhancing effect in the germa-
nium chemistry described herein. Preliminary enantiose-
lectivity testing of (1R,2S,5R)-menthyldiphenylgermane
ꢀ
1 ꢀ1
1
9
1
-hexenyl radical with a value for log(A/M
.95± 0.22 and an activation energy (Ea) of
s ) of
ꢀ1 34
9.08± 1.52kJmol
.
The data can be compared with
Table 2. Kinetic parameters for the reactions of primary alkyl radicals
with (1R,2S,5R)-menthyldiphenylgermane 9 and tributylgermane in
tert-butylbenzene (60–120ꢀC) and comparative data for triphenylger-
mane
9
7
and its enantiomer ent-9 against selected substrates
, 15–18 was preformed in toluene at ꢀ78ꢀC according
ꢀ1
ꢀ1a
ꢀ1 a
b
Germane
logA/M
s
E
a
(kJmol
)
k (80ꢀC)
·10 M
H
to our previously-published protocol, the results of
which are listed in Table 3. The substrates chosen for
6
ꢀ1 ꢀ1
s )
(
6
MenPh
2
GeH 9 11.13± 0.44
8.60± 0.46
34.57± 3.05
20.70± 3.21
1.02
0.34
0.30
this investigation are representative of those previously
used by us and serve as appropriate benchmarks.
Bu
3
GeH
c
c
8.44± 0.47
19.7± 1.86
d
Ph
3
GeH
––
––
3.8
a
Inspection of Table 3 reveals that chiral germanes ap-
pear to behave in a very similar manner to their tin
counterparts in that poor enantioselectivities are ob-
served in the absence of a Lewis acid, in this case MgBr2.
In every entry, the addition of the magnesium salt had a
Error limits are expressed to 95% confidence but include random and
not systematic variations.
b
c
Calculated from the Arrhenius parameters.
Ref. 16.
Ref. 33.
d

L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
2551
O
O
O
Finally, the low to moderate reaction yields and conver-
sions deserve mention. Despite our efforts to improve the
yield through modification of reaction conditions, we
were never able to achieve outcomes that exceeded about
50%. We speculate that poor radical chain propagation
due to slow germane hydrogen transfer rate constants
at low temperature (compared to their tin counterparts)
HN
Br
CF3
HN
Br
CF3
OBn
OMe
O
1
5
16
3
6
may be partly responsible for this observation. Indeed,
low-temperature reductions involving 9 required re-
peated re-initiation. It should be noted that 8 provides
yields up to 97%, but is some 15 times more reactive that
CH3
Br
Br
OEt
Ph
O
O
7
ꢀ1 ꢀ1
8
9
, with a reported rate constant of about 1.4·10 M
s
1
at 80ꢀC for reaction with primary alkyl radicals.
1
7
18
dramatic effect on the observed ee. For example, entry 1
demonstrates that in the absence of MgBr , 2-bromon-
3. Experimental
2
aproxen ester 7 affords the reduced product, naproxen
ester 6 in 30% conversion with only 10% ee. While the
addition of MgBr appears not to improve the reaction
Substrates 7, 15–18 and reduction standards were pre-
6
pared following our previously published protocols.
H and C NMR spectra were obtained using a Jeol
1
13
2
yield, it does have the same dramatic increase on the
reactionꢁs ee as has been observed previously in chiral
stannane chemistry, with 6 being produced in 99% ee.
GX 270, Varian 300MHz Unity Plus, or Jeol Eclipse
Plus 400 NMR spectrometer. Chemical shifts (d) are
given in ppm and are referenced against tetramethylsi-
lane (TMS). Gas chromatographic analyses were per-
formed using a chiral trifluoroacetylated c-cyclodextrin
6
Importantly, as is clearly evident in Table 3, the two
enantiomeric germanes 9 and ent-9 provided opposite
stereochemical outcomes for each substrate, as expected.
The data obtained in this work can be compared with
those of Curran and co-workers, who reported ees of
up to 42% using chiral germane 8, albeit in the absence
(
ChiraldexTM G–TA, 30m·0.25mm) capillary column
purchased from Alltech. HPLC analyses were carried
out using a Regis (S,S) Whelk-O 1 (25cm·4.5mm ID)
column.
1
8
of Lewis acid additives. While the selectivities reported
herein are high in the presence of MgBr , the ees re-
2
(1R,2S,5R)-(ꢀ)-Menthylgermanium trichloride 10: A
solution of freshly-prepared (1R,2S,5R)-menthylmagne-
sium chloride, prepared from (1R,2S,5R)-menthyl chlo-
ride (8.15g, 46.6mmol) and magnesium (1.25g,
ported by Curran exceed those obtained in this study
in the absence of Lewis acid additives, suggesting that
the chiral environment provided by structures such as
8 is more effective in imparting chiral recognition than
that provided by 9.
5
1.3mmol) in THF (50mL) was added via cannula to
a cooled (ꢀ20ꢀC) stirred solution of germanium tetra-
chloride (10.0g, 46.6mmol) in ether (100mL). The reac-
tion mixture was allowed to warm to room temperature
and stirred overnight. After removal of the precipitate
by filtration and removal of the solvent in vacuo, distil-
lation (94ꢀC, 5Pa) gave the title compound as a colour-
Table 3. Enantioselectivities observed for reductions of substrates 7,
15–18 with (1R,2S,5R)-menthyldiphenyl-germane 9 and its enantiomer
ent-9 in toluene at ꢀ78ꢀC
Entry Substrate Lewis
b
Germane % Ee % Yield Config
a
25
1
acid
less oil (7.4g, 50% yield). ½aꢁ ¼ ꢀ51:3 (c 1, CHCl ). H
D
3
NMR (299.98MHz, CDCl ): d 0.87 (d, 3H, CH ), 0.96
1
2
3
4
7
15
16
None
MgBr
None
MgBr
9
10
99
5
30
30
20
15
S
S
R
R
3 3
(d, 3H, CH ), 0.99 (d, 3H, CH ), 1.00–1.26 (m, 2H),
1.14–1.24 (m, 1H), 1.62–1.86 (m, 4H), 1.96–2.08 (m,
1H), 2.13–2.23 (m, 1H), 2.38–2.50 (m, 1H); C{ H}
NMR (75.44MHz, CDCl ): d 15.33 (CH ), 21.51
2
2
3
3
ent-9
c
13
1
99
3
3
5
6
7
8
None
MgBr
None
MgBr
9
10
80
5
20
20
15
20
S
S
R
R
(
(
(
CH ), 22.30 (CH ), 25.30 (CH ), 31.36 (CH), 34.07
CH), 34.24 (CH ), 36.34 (CH ), 44.70 (CH), 50.54
3 3 2
2
2
2
2
ent-9
+
CH); ESMS (+ve) 341.0 (M+Na) . Anal. Calcd for
c
90
C H GeCl (318.23): C 37.74, H 6.02; found: C
1
3
0
19
3
7.80%, H 6.00%.
9
1
1
None
MgBr
MgBr
9
0
99
99
20
40
50
––
S
0
1
2
2
(
1R,2S,5R)-(ꢀ)-Menthyltriphenylgermane 11: A solu-
ent-9
R
tion of phenylmagnesium bromide, was prepared from
bromobenzene (29.6g, 188.5mmol) and magnesium
(4.58g, 188.5mmol) in THF (200mL). The solution
was filtered into a second flask, through a sintered glass
frit under nitrogen, while hot, to remove unreacted mag-
nesium. (1R,2S,5R)-(ꢀ)-Menthylgermanium trichloride
10 (10.0 g, 31.4mmol) in THF (20mL) was added
dropwise at room temperature to the filtered solution.
The reaction mixture was stirred overnight before being
1
1
2
3
17
18
None
MgBr
9
10
99
30
30
S
S
2
2
1
1
4
5
None
MgBr
9
10
99
35
30
S
S
a
b
c
See Ref. 35.
GC conversion.
Isolated yield.

2552
L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
carefully quenched with water. The solvent was removed
in vacuo and ether (200mL) and water (100mL) added.
After filtration and separation, the organic layer was
dried over Na SO , and the solvent removed in vacuo.
(1S,2R,5S)-(+)-Menthylgermanium trichloride ent-10
was prepared as described above from (1S,2R,5S)-men-
25
thyl chloride in 60% yield). ½aꢁ ¼ þ45:9 (c 1, CHCl3).
D
2
4
Crystallization from hexane at 4ꢀC gave the title com-
pound as a colourless solid (13.0g, 93% yield). Mp
(1S,2R,5S)-(+)-Menthyltriphenylgermane ent-11 was
prepared as described above from (1S,2R,5S)-(+)-men-
thylgermanium trichloride ent-10 in 90% yield. Mp
25
1
9
7–98ꢀC. ½aꢁ ¼ ꢀ34:8 (c 1, CHCl ). H NMR
D
3
20
(
299.98MHz, CDCl ): d 0.77 (d, 3H, CH ), 0.82 (d,
101–102ꢀC. ½aꢁ ¼ þ37:1 (c 1, CHCl3).
3 3
D
3
1
1
9
H, CH ), 0.97 (d, 3H, CH ), 1.00–1.15 (m, 2H), 1.20–
.35 (m, 2H), 1.40–1.55 (m, 1H), 1.55–1.65 (m, 1H),
.75–2.05 (m, 3H), 2.10–2.30 (m, 1H), 7.40–7.60 (m,
3 3
(1S,2R,5S)-(+)-Menthyldiphenylgermanium
bromide
ent-12 was prepared as described above from
(1S,2R,5S)-(+)-menthyltriphenylgermane ent-11 in
1
3
1
C{ H} NMR
H, Ph), 7.65–7.85 (m, 6H, Ph);
21
(
75.44MHz, CDCl ): d 16.36 (CH ), 21.58 (CH ),
quantitative yield as a viscous oil. ½aꢁ ¼ þ43:5 (c 1,
3
3
3
D
2
3
1
2.65 (CH ), 26.37 (CH ), 30.81 (CH), 31.47 (CH),
4.66 (CH), 35.21 (CH ), 39.68 (CH ), 45.21 (CH),
CHCl ).
3
3
2
2
2
27.92 (Ph ), 128.37 (Ph ), 135.29 (Ph ), 137.87 (Ph );
m
(1S,2R,5S)-(+)-Menthyldiphenylgermane ent-9 was
prepared as described above from (1S,2R,5S)-(+)-men-
thyldiphenylgermanium bromide ent-12 in 88% yield.
p
o
i
Anal. Calcd for C H Ge (443.19): C 75.88%, H
34
2
8
7
.73%; found: C 75.80%, H 7.81%.
2
D
1
½
aꢁ ¼ þ30:0 (c 1, toluene).
(
1
1R,2S,5R)-(ꢀ)-Menthyldiphenylgermanium bromide
2: Bromine (5.06g, 31.63mmol) was added dropwise
Typical kinetic experiment: Following our previously de-
4
3
to a solution of (1R,2S,5R)-(ꢀ)-menthyltriphenylger-
scribed procedure, an aliquot (100lL) of a standard
mane 11 (14.02g, 31.63mmol) in dibromoethane
(
solution (0.05–0.15M) of germane (9 or Bu GeH) in
3
50mL). The solution was stirred overnight at reflux,
tert-butylbenzene was placed in a small Pyrex tube, at
which point 1-bromo-5-hexene (ca. 0.1equiv) and AIBN
(ca. 1 crystal) were added and the solution degassed by
the usual freeze-thaw technique, before being sealed
under vacuum. After being thermolyzed in an oil bath
at the required temperature, the solution was analyzed
by GC.
after which the solvent removed in vacuo to give the title
bromide as a pale yellow oil, which solidified on stand-
ing and used without further purification (13.8g, 98%).
2
0
1
Mp 51–52ꢀC. ½aꢁ ¼ ꢀ44:9 (c 1, CHCl ). H NMR
D
3
(
270.17MHz, CDCl ): d 0.64 (d, 3H, CH ), 0.79 (d,
3 3
3
1
2
H, CH ), 0.89 (d, 3H, CH ), 0.90–1.30 (m, 3H), 1.30–
.50 (m, 1H), 1.50–1.65 (m, 1H), 1.65–1.90 (m, 3H),
.00–2.25 (m, 2H), 7.30–7.80 (m, 10H, Ph); C{ H}
3 3
13
1
Standard procedure for small-scale low-temperature ger-
mane reductions: Following our previously described
NMR (67.94MHz, CDCl ): d 15.68 (CH ), 21.51
3
3
6
(
(
(
(
CH ), 22.48 (CH ), 25.83 (CH ), 30.96 (CH), 34.33
CH), 34.89 (CH ), 36.16 (CH), 38.51 (CH ), 45.13
procedure, a flask fitted with a septum was charged
3
3
2
with a solution of the required bromide (0.1mmol)
and internal standard (octane or decane, 0.1mmol) in
toluene (0.5mL) and 9-BBN (a few crystals) added.
The solution was cooled to the required temperature,
the flask purged with nitrogen and the required chiral
germane (9 or ent-9) (0.11mmol) in toluene (0.5mL)
added. The reaction mixture was stirred at the required
temperature for 8h, with an additional amount of tri-
ethylborane (0.05mL of a 1M solution in THF) added
every 2h. The solution was warmed to room tempera-
ture and analyzed directly by GC or HPLC.
2
2
CH), 128.26/128.34 (Ph ), 129.68 (Ph ), 133.70/133.81
m
p
Ph ), 137.14/136.04 (Ph ).
i
o
(
1R,2S,5R)-(ꢀ)-Menthyldiphenylgermane 9: A solution
of crude (1R,2S,5R)-menthyldiphenylgermanium bro-
mide 12 (10.0g, 22.4mmol) in ether (50mL) was added
dropwise to a suspension of lithium aluminiumhydride
(
0.43g, 11.2mmol) in ether (50mL). The reaction mix-
ture was stirred at room temperature for 1h and then
quenched with water (50mL) (ice cooling bath). After
filtration, the organic layer was collected and the aque-
ous layer extracted with ether (50mL). The combined
organic layers were dried over Na SO and the solvent
Reduction of 2-bromonaproxen ethyl ester 7 with
(1S,2R,5S)-menthyldiphenylgermane ent-9: Magnesium
bromide etherate (MgBr ÆEt O) (0.36g, 1.38mmol) was
2
4
removed in vacuo. Distillation (140ꢀC, 1Pa) (K u¨ gel-
rohr) gave the title germane as a colourless oil (7.1g,
2
2
added to dry toluene (3mL) and the mixture allowed
to stir for 30min under N . Bromoester 7 (0.246g,
2
2
1
8
7% yield). ½aꢁ ¼ ꢀ32:2 (c 1, toluene). H NMR
D
2
(
CH ), 0.78 (d, 3H, CH ), 0.82–1.04 (m, 2H), 1.08–1.24
299.98MHz, C D ): d 0.73 (d, 3H, CH ), 0.75 (d, 3H,
0.690mmol) in dry toluene (0.2mL) was added slowly
to the reaction mixture, which was allowed to stir at rt
for a further 10min prior to cooling to ꢀ78ꢀC. After
stirring at ꢀ78ꢀC for a further 45min, (1R,2S,5R)-ment-
hyldiphenylgermane ent-9 (0.36g, 0.715mmol) in tolu-
ene (3mL) was slowly added, followed by
triethylborane (0.2mL of 1M solution in THF) and oxy-
gen introduced. The reaction mixture was stirred at this
temperature for a further 8h. Additional triethylborane
(0.2mL of 1M solution in THF) was added to the reac-
tion mixture every 2h. After 8h there was no further
change as evidenced by TLC. The mixture was quenched
6
6
3
3
3
(
m, 2H), 1.32–1.44 (m, 1H), 1.50–1.76 (m, 3H), 1.88–
1
7
.98 (m, 1H), 1.98–2.14 (m, 1H), 5.31 (s, 1H, GeH),
.08–7.20 (m, 6H, Ph), 7.52–7.60 (m, 4H, Ph); C{ H}
13
1
NMR (75.44MHz, C D ): d 15.82 (CH ), 21.91 (CH ),
6
6
3
3
2
3
1
2.81 (CH ), 26.32 (CH ), 31.11 (CH), 31.90 (CH),
4.85 (CH), 35.57 (CH ), 39.86 (CH ), 45.84 (CH),
3 2
2
2
28.56/128.61 (Ph ), 129.04/129.07 (Ph ), 135.58/135.71
m
p
(
Ph ), 136.85/136.93 (Ph ). Anal. Calcd for C H Ge
o i 22 30
(
8
367.08): C 71.98%, H 8.24%; found: C 71.93%, H
.39%.

L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
2553
with water (2mL) and extracted with ether (2·). The or-
References and notes
ganic layer was dried over MgSO and excess solvent re-
4
moved in vacuo to afford the crude product as light
yellow oil. Further purification of the product (flash
chromatography, 96:4 hexane/ethyl acetate) yielded benz-
1. Carland, M. W.; Schiesser, C. H. In The Chemistry of
Organic Germanium, Tin and Lead Compounds; Rappo-
port, Z., Ed.; Wiley: Chichester, UK, 2002; Vol. 2, nd
references cited therein.
2. Giese, B. Radicals in Organic Synthesis: Formation of
Carbon–Carbon Bonds; Pergamon: Oxford, 1986.
yl (R)-naproxen ethyl ester as a colourless oil (0.027g,
1
1
m), 4.1 (2H, m), 3.9 (3H, s), 3.8 (1H, q), 1.6 (3H, d),
5% yield, 99% ee). H (NMR) CDCl : d 7.8–7.1 (6H,
3
3
. Renaud, P.; Sibi, M. P. In Radicals in Organic Synthesis;
Wiley–VCH: Weinheim, 1989; Vols. 1 and 2; Perkins, M.
J.; Ellis-Horwood: New York, 2001.
14
1.4 (3H, t). ½aꢁ ¼ ꢀ32:6 (c 0.12, CHCl ). The sample
was identical to that prepared previously.
D
3
6
4
. or excellent reviews, see: (a) Curran, D. P.; Porter, N. A.;
Giese, B. Stereochemistry of Radical Reactions; VCH
Weinheim, 1995; (b) Sibi, M.; Porter, N. A. Acc. Chem.
Res. 1999, 32, 163.
Reduction of benzyl N-trifluoroacetyl-2-bromo-tert-leuci-
nate 15 with (1S,2R,5S)-menthyldiphenylgermane ent-9:
Following the protocol described above, bromoester
5
. Dakternieks, D.; Dunn, K.; Perchyonok, V. T.; Schiesser,
C. H. Chem. Commun. 1999, 1665.
6. Dakternieks, D.; Perchyonok, V. T.; Schiesser, C. H.
Tetrahedron: Asymmetry 2003, 14, 3057.
. Dakternieks, D.; Dunn, K.; Henry, D. J.; Schiesser, C. H.;
Tiekink, E. R. T. Organometallics 1999, 18, 3342.
1
as a colourless oil (20%) with identical properties to
5 afforded benzyl (R)-N-trifluoroacetyl-tert-leucinate
6
1
those prepared previously. H NMR CDCl : d 7.2
3
(
8
5H, m), 7.6 (1H, br s), 5.3 (2H, m), 4.5 (1H, d, J
Hz), 1.0 (9H, s). ½aꢁD ¼ þ8:3 (c 0.4, CHCl3).
7
8
14
. Dakternieks, D.; Dunn, K.; Schiesser, C. H.
Tiekink, E. R. T. J. Organomet. Chem. 2000, 605, 209.
Crystallographic studies: Single crystals of (1R,2S,5R)-
(
ꢀ)-menthyltriphenylgermane 11 and its enantiomer
9. (a) Chatgilialoglu, C.; Griller, D.; Lesage, M. J. Org.
Chem. 1988, 53, 3641; (b) Ballestri, M.; Chatgilialoglu, C.;
Clark, K. B.; Griller, D.; Giese, B.; Kopping, B. J. Org.
Chem. 1991, 56, 678.
ent-11 suitable for X-ray crystallography were obtained
from hexane at room temperature. Crystal data and
structure solutions at T=293(2)K: 11 C H Ge,
2
8
34
1
0. Crich, D.; Sun, S. J. Org. Chem. 1996, 61, 7200.
M =443.14, orthorhombic, P2 2 2 , a=9.3345(6), b=
r
1 1 1
˚ ˚
3
2.5991(8), c=20.6109(12)A, V=2424.0(3)A , Z=4,
11. (a) Chatgilialoglu, C. Acc. Chem. Res. 1992, 25, 188, nd
references cited therein; (b) Chatgilialoglu, C. Chem. Rev.
1
3
˚
D =1.214Mg/m , F(000)=936, k(MoKa)=0.71073A,
x
1
995, 95, 1229, nd references cited therein.
ꢀ
1
l=1.274mm
orthorhombic, P2 2 2 , a=9.3103(7), b=12.5683(9),
and ent-11 C H Ge, M =443.14,
28 34 r
12. (a) Chatgilialoglu, C.; Griller, D.; Lesage, M. J. Org.
Chem. 1988, 53, 3641; (b) Ballestri, M.; Chatgilialoglu, C.;
Clark, K. B.; Griller, D.; Giese, B.; Kopping, B. J. Org.
Chem. 1991, 56, 678.
13. Chatgilialoglu, C.; Schiesser, C. H. In The Chemistry of
Organic Silicon Compounds; Rappoport, Z., Apeloig, Y.,
Eds.; Wiley: Chichester, UK, 2001; Vol. 3, nd references
cited therein.
1
1 1
3
V=2405.8(3)A ,
˚
c=20.5599(15)A,
˚
Z=4,
D =
x
3
.223Mg/m ,
˚
1
F
(000)=936, k(MoKa)=0.71073A,
ꢀ
1
l=1.284mm . The data were collected to a maximum
h=27.49ꢀ 11 and 27.49ꢀ ent-11 with 3 sets at different j-
angles and 362 frames via x-rotation (D/x=1ꢀ) at two
times 10s per frame on a Nonius Kappa CCD diffrac-
tometer with a completeness of 99.8% (11) and 99.8%
ent-11 (hmax). The structures were solved by direct meth-
1
4. Ballestri, M.; Chatgilialoglu, C.; Seconi, G. J. Organomet.
Chem. 1991, 408, C1.
1
5. Johnston, L. J.; Lusztyk, J.; Wayner, D. D. M.; Abey-
wickrema, A. N.; Beckwith, A. L. J.; Scaiano, J. C.;
Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 4594.
3
7
ods using SHELXS-97 and refined by full-matrix least-
2
squares calculations using all measured F data and
3
8
SHELXL-97. All non-H atoms were refined anisotropi-
cally. The H atoms were placed in geometrically calcu-
lated positions using a riding model. Atomic scattering
factors for neutral atoms and real and imaginary disper-
sion terms were taken from International Tables for
16. (a) Lusztyk, J.; Maillard, B.; Lindsay, D. A.; Ingold, K. U.
J. Am. Chem. Soc. 1983, 105, 3578; (b) Lusztyk, J.;
Maillard, B.; Deycard, S.; Lindsay, D. A.; Ingold, K. U.
J. Org. Chem. 1987, 52, 3509.
7. Chatgilialoglu, C. M. Adv. Organomet. Chem. 1999, 44,
67.
18. Gualtieri, G.; Geib, S. J.; Curran, D. P. J. Org. Chem.
003, 68, 5013.
9. Spivey, A. C.; Diaper, C. M. Sci. Synth. 2003, 5, 159.
20. Yarosh, O. G.; Yarosh, N. O.; Albanov, A. I.; Voronkov,
M. G. Russ. J. Gen. Chem. 2002, 72, 1901.
1
3
9
X-ray Crystallography. The figures were created by
4
0
DIAMOND. R =0.0393 for 5161 [I>2r(I)] and
1
2
wR =0.0857 for 5533 independent reflections 11 and
R =0.0499 for 4814 [I>2r(I)] and wR =0.0982 for
2
1
1
2
5
503 independent reflections ent-11. The max. and
min. residual electron densities were 0.598/ꢀ0.226
21. Dakkouri, M.; Kehrer, H. Chem. Ber. 1983, 116, 2041.
22. Dakternieks, D.; Duthie, A.; Schiesser, C. H.,
unpublished.
ꢀ
3
ꢀ3
˚
˚
eA for 11 and 0.631/ꢀ0.449eA for ent-11. Crystal-
lographic data (excluding structure factors) for the
structures in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplemen-
tary publication numbers CCDC 240940 and 24094.
2
2
3. Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317.
4. Dakternieks, D.; Henry, D. J.; Schiesser, C. H. J. Phys.
Org. Chem. 1999, 12, 233.
2
5. Walling, C.; Cooley, J. H.; Ponaras, A. A.; Racah, E. J.
J. Am. Chem. Soc. 1966, 88, 5361.
2
2
6. Walling, C.; Cioffari, A. J. Am. Chem. Soc. 1972, 94, 6059.
7. Beckwith, A. L. J.; Moad, G. J. Chem. Soc., Chem.
Commun. 1974, 472.
Acknowledgements
Financial support from the Australian Research Council
and Chirogen Pty. Ltd is gratefully acknowledged.
28. Beckwith, A. L. J.; Lawrence, T. J. Chem. Soc., Perkin
Trans. 2 1979, 1535.

2554
L. Zeng et al. / Tetrahedron: Asymmetry 15 (2004) 2547–2554
2
3
3
3
3
3
9. Beckwith, A. L. J.; Moad, G.. J. Chem. Soc., Perkin Trans.
2
0. Lal, D.; Griller, D.; Husband, S.; Ingold, K. U. J. Am.
Chem. Soc. 1974, 96, 6355.
1. Schmid, P.; Griller, D.; Ingold, K. U. Int. J. Chem. Kinet.
35. While the use of more than 1equiv of fresh Lewis acid
1980, 1083.
provided no increase in observed enantioselectivity, often
2equiv were added when older stock of Lewis acid were
employed.
36. The Arrhenius equation for 9 provides a rate constant of
1
ꢀ1 ꢀ1
s
1979, 11, 333.
6.6·10 M
at ꢀ78ꢀC.
2. Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J. C. J. Am.
Chem. Soc. 1981, 103, 7739.
´
3. Chatgilialoglu, C.; Ballestri, M.; Escudıe, J.; Pailhous, I.
Organometallics 1999, 18, 2395.
4. Zeng, L.; Perchyonok, V. T.; Schiesser, C. H. Tetrahedron:
Asymmetry 2004, 15, 995.
37. Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
38. Sheldrick, G. M. University of G o¨ ttingen, 1997.
39. International Tables for Crystallography; Kluwer Aca-
demic: Dordrecht, 1992; Vol. C.
40. DIAMOND V2.1d, Crystal Impact, K. Brandenburg &
M. Berndt GbR, http://www.crystalimpact.de.