A new family of chelating diphosphines with a transition metal stereocenter in the backbone: Novel applications of "chiral-at-rhenium" complexes in rhodium-catalyzed enantioselective alkene hydrogenations

Article abstract of DOI:10.1002/1521-3765(20010504)7:9<2015::AID-CHEM2015>3.0.CO;2-H

The title compounds are accessed by sequences starting with racemic and enantiomerically pure [(η⁵-C₅H₅)Re(NO)(PPh₃) (CH₃)]. Reactions with chlorobenzene/HBF₄, PPh₂H, and tBuOK give the phosphido complex [(η⁵-C₅H₅)Re(NO) (PPh₃)(PPh₂)] (3). Reactions with Ph₃C⁺BF₄^-,PPh₂H, and tBuOK give the methylene homologue [(η⁵-C₅H₅)Re(NO)(PPh₃) (CH₂PPh₂)] (9). Treatment of 3 or 9 with nBuLi or tBuLi and then PPh₂Cl gives the diphosphido systems [(η⁵-C₅H₄PPh₂)Re(NO) (PPh₃)-((CH₂)_nPPh₂)] (n = 0/1, 5/11). Reactions of 5 and 11 with [Rh(NBD)Cl]₂/AgPF₆ (NBD = norbornadiene) give the rhenium/rhodium chelate complexes [(η⁵-C₅H₄-PPh₂)Re(NO) (PPh₃)((μ-CH₂)_nPPh ₂)Rh-(NBD)]⁺PF₆^- (n = 0/1, 6⁺/12⁺ PF₆^-; 30-32% overall from commercial Re₂(CO)₁₀). The crystal structures of 6⁺ PF₆^- and 12⁺ PF₆^- are compared to those of 3 and 9, and other rhodium complexes of chelating bis(diphenylphosphines). The chiral pockets defined by the PPh₂ groups show unusual features. Four alkenes of the type (Z)-RCH=C(NHCOCH₃)CO₂R′ are treated with H₂ (1 atm) and (R)-6⁺ PF₆^- or (S)-12⁺ PF₆^- (0.5 mol%) in THF at room temperature. Protected amino acids are obtained in 70-98% yields and 93-82 % ee [(R)-6⁺ PF₆^-] or 72-60% ee [(S)-12⁺ PF_6- ]. Pressure and temperature effects are defined, and turnover numbers of > 1600 are realized.

Full text of DOI:10.1002/1521-3765(20010504)7:9<2015::AID-CHEM2015>3.0.CO;2-H

FULL PAPER
A New Family of Chelating Diphosphines with a Transition Metal
Stereocenter in the Backbone: Novel Applications of ªChiral-at-Rheniumº
Complexes in Rhodium-Catalyzed Enantioselective Alkene Hydrogenations
[
a]
[b]
[b]
[a]
Klemenz Kromm, Bill D. Zwick, Oliver Meyer, Frank Hampel, and
John A. Gladysz*[
a]
Abstract: The title compounds are ac-
of 5 and 11 with [Rh(NBD)Cl] /AgPF6
those of 3 and 9, and other rhodium
complexes of chelating bis(diphenyl-
phosphines). The chiral pockets defined
2
cessed by sequences starting with race-
(NBD ˆ norbornadiene) give the rheni-
5
5
mic and enantiomerically pure [(h -
um/rhodium chelate complexes [(h -C H -
5
4
C H )Re(NO)(PPh )(CH )]. Reactions
PPh )Re(NO)(PPh )((m-CH ) PPh )Rh-
by the PPh groups show unusual fea-
5
5
3
3
2
3
2
‡
n
2
2
‡
�
‡
�
with chlorobenzene/HBF , PPh H, and
(NBD)] PF6 (n ˆ 0/1, 6 /12 PF6
;
tures. Four alkenes of the type (Z)-
RCHˆC(NHCOCH )CO R' are treated
4
2
tBuOK give the phosphido complex
30 ± 32% overall from commercial
Re (CO) ). The crystal structures of
3
2
5
‡
�
6
[
(h -C H )Re(NO)(PPh )(PPh )]
(3).
with H (1 atm) and (R)-6 PF or (S)-
2
5
5
3
2
2
10
‡
�
‡
�
‡
�
‡
�
Reactions with Ph C BF , PPh H, and
6 PF and 12 PF6 are compared to
12 PF6 (0.5 mol%) in THF at room
temperature. Protected amino acids are
obtained in 70 ± 98% yields and
93 ± 82% ee [(R)-6 PF ] or 72 ± 60%
ee [(S)-12 PF ]. Pressure and temper-
3
4
2
6
tBuOK give the methylene homologue
5
[
(h -C H )Re(NO)(PPh )(CH PPh )] (9).
5
5
3
2
2
Keywords: asymmetric hydrogena-
‡
�
Treatment of 3 or 9 with nBuLi or tBuLi
6
tion
´ amino acids ´ catalyst ´
‡
�
and then PPh Cl gives the diphosphido
2
6
conformation analysis ´ heterobime-
tallic
5
systems [(h -C H PPh )Re(NO)(PPh )-
ature effects are defined, and turnover
numbers of >1600 are realized.
5
4
2
3
(
(CH ) PPh )] (n ˆ 0/1, 5/11). Reactions
2
n
2
Introduction
M(A)(B)(C)(D)Ðmight also provide efficient and perhaps
superior stereogenesis. We thought that the chelate backbone
would be a particularly favorable position for such a group, as
represented schematically by B in Scheme 1.
The design of new enantioselective catalysts is both an art and
a science. For inspiration, chemists have considered virtually
every type of chiral building block available in non-racemic
Such ligands can be viewed as relatives of classical chiral
[
1, 2]
[5]
form.
For example, the use of metal catalysts featuring
chelates such as DIOP and chiraphos, with the carbon
ferrocene-based chelating ligands with ªplanar chiralityº has
stereocenters replaced by a metal stereocenter. They offer a
number of potential advantages. First, metal-based stereo-
centers constitute extremely flexible diversity elements.
Second, steric properties can be fine-tuned in numerous ways.
Third, electronic effects of metals are often transmitted over
considerable distances,[ and could be employed to either
stabilize or (hemi)labilize a chelate. One concern might be
whether metal-containing ligands will be as robust as carbon
analogues. However, ferrocenes are sensitive towards electro-
philes and oxidizing agents, but chelating ligands of the types
A/A' yield metal catalysts that give very high turnover
[
3, 4]
grown rapidly over the last decade.
Many have proved
spectacularly successful, and two representative ligand classes
are illustrated in Scheme 1 (A, A'). This led us to speculate
that chelating ligands that incorporate a chiral metal centerÐ
for example, a non-planar spectator moiety of general formula
6]
[
a] Prof. Dr. J. A. Gladysz, Dipl.-Chem. K. Kromm, Dr. F. Hampel
Institut für Organische Chemie
Universität Erlangen-Nürnberg
Henkestrasse 42, 91054 Erlangen (Germany)
E-mail: gladysz@organik.uni-erlangen.de
[3a,e]
numbers and are applied in industrial processes.
The field of ªchiral-at-metalº complexes has been pioneered
by H. Brunner.[ Several enantioselective catalysts bearing a
M(A)(B)(C)(D) substructure or active site have been re-
[
b] Dr. B. D. Zwick, Dr. O. Meyer
7]
Department of Chemistry, University of Utah
Salt Lake City, Utah 84112 (USA)
ported, but all examples to date contain additional carbon
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry/ or from the author.
stereocenters.[
2, 7, 8]
We have conducted extensive studies of
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2015 $ 17.50+.50/0
2015

J. A. Gladysz et al.
FULL PAPER
R
There is an extensive literature of enantioselective catalysts
R
[1±5, 16]
bearing chelating diphosphine ligands.
In view of the
:
E
A
E:
E:
many benchmarks available, coupled with numerous unsolved
problems relating to enantiomeric excesses, rates, and yields,
we began our efforts in this area. In the narrative below, we
describe new enantioselective alkene hydrogenation catalysts
E:
Fe
Fe
M
C
D
E:
B
E:
R
that provide a convincing proof-of-concept. Additional appli-
A
A'
B
[17]
cations of our ligand systems will reported elsewhere.
A
planar chirality
mixed planar and
tetrahedral chirality
chiral-at-metal
[
18]
small portion of this work has been communicated.
25°
135°
60°
60°
75°
5° 45°PPh3
Results
1
35°
1
75°
ON
ON
Re
Non-racemic five-membered chelates: The non-racemic
°
125°
90
5
ON
PPh3
X
0° PPh3
methyl complex (S)-[(h -C
H
5
)Re(NO)(PPh
5
3
)(CH
3
)], [(S)-
4
9
[19, 20]
1
],
was prepared from commercial Re (CO) in a routine
series of steps as previously described. An enantiomeric
purity of >99% ee was verified by HPLC.[ As shown in
2
9]
10
[
0°
9
21]
C
D
E
Scheme 2, (S)-1 and etheral HBF were combined in chlor-
4
obenzene at � 418C. This generates a substitution-labile,
idealized electronic and structural features of title compounds
Scheme 1. Chiral organometallic scaffolds for chelating ligands.
+
BF4-
PPh3
a)
b)
Re
Re
ON
PPh3
PPh3
PPh3
ON
5
chiral rhenium complexes of the formula [(h -C H )Re-
CH3
5
5
PPh2H
(
NO)(PPh )(X)] (see C ± E in Scheme 1). These are easily
obtained in enantiomerically pure form and with only a few
3
+
-
(
S)-1
(R)-2 BF
4
[
9]
exceptions (X ˆ OR, NR ) are configurationally robust at
2
c)
[
10]
ambient temperature. However, most of the enantioselec-
tive organic transformations developed to date have been
[
11]
Li
stoichiometric in rhenium. In view of the flexibility with
which this template can be elaboratedÐvirtually any type of
X group is possible, other phosphorus donor ligands can be
employed, and substituted cyclopentadienyl ligands can be
introducedÐwe set out to incorporate it into chelate ligands
of the type B (Scheme 1).
Some key properties of these compounds deserve emphasis
at the outset. First, the sixteen-valence-electron fragment [(h -
C H )Re(NO)(PPh )] is both a Lewis acid and a strong p
base, with the d-orbital HOMO shown in C. This is a key
d)
Re
Re
ON
ON
PPh3
PPh2
PPh2
(R)-4
(R)-3
e)
5
‡
+
5
5
3
Ph2P
Ph2P
Rh
f)
PF6-
NO
Re
Re
determinant of conformation in adducts of unsaturated
ON
[
12]
ligands (attractive interactions) and saturated ligands with
lone pairs on the ligating atom (repulsive interactions).[
Second, such complexes are formally octahedral, with the
cyclopentadienyl ligand occupying three coordination sites as
shown in D. Thus, there are small but sometimes important
differences in bond and torsion angle relationships as
compared to carbon stereocenters, some of which are evident
in E. Third, the basicity and nucleophilicity of any lone pair on
the ligating atom X is enhanced relative to organic ana-
This has been most clearly documented with
phosphido complexes [(h -C H )Re(NO)(PPh )(PR )], which
are alkylated by CH Cl at room temperature, and has its
main origin in the rhenium/lone pair repulsive interactions
noted above. Fourth, this effect persists with ligands of the
P
Ph2
PPh2
PPh3
13]
(S)-6+ PF6-
(R)-5
Scheme 2. Synthesis of the non-racemic five-membered chelating diphos-
[
19]
phine. , � 418C; b) PPh
58C; d) nBuLi, THF, � 788C; e) PPh
AgPF , THF, 208C.
a) Chlorobenzene, HBF
4
2
H; c) tBuOK, THF,
Cl, THF, � 788C; f) [Rh(NBD)Cl]
2
2
2
,
6
cationic chlorobenzene adduct that serves as a functional
[
13, 14]
5
logues.
equivalent of the chiral Lewis acid [(h -C H )Re(NO)-
5
5
5
‡
[22]
(PPh )] . Addition of PPh H gave the diphenylphosphine
3 2
5
5
3
2
[
13]
5
‡
�
complex (R)-[(h -C H )Re(NO)(PPh )(PPh H)] BF , [(R)-
5 5 3 2
2
2
4
‡
�
2 BF ], in 89% yield after workup. The racemic tosylate salt
4
‡
�
[13a, 23]
2 OTs has been prepared by a related route.
Reaction
‡
�
type -CH X', as most clearly demonstrated for sulfur-contain-
of (R)-2 BF and tBuOK as previously described for
2 OTs gave the diphenylphosphido complex (R)-[(h -
2
4
[
15]
‡
�
5
ing species.
2016
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2016 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

ªChiral-at-Rheniumº Complexes
2015±2027
C H )Re(NO)(PPh )(PPh )], [(R)-3], in 99% yield after
and workup gave the target diphenylphosphidocyclopenta-
5
5
3
2
5
workup.
dienyl complex (R)-[(h -C H PPh )Re(NO)(PPh )(PPh )],
5
4
2
3
2
‡
�
Complexes (R)-2 BF and (R)-3 exhibited good thermal
[(R)-5], in 89% yield as a spectroscopically pure red foam.
Crystallization of (R)-5 from benzene/hexane gave red
prisms of a benzene hemisolvate. Racemic 5 was prepared by
4
stabilities. However, the latter was very air sensitive in
solution and the solid state, analogous to the racemate, which
[
13a]
1
13
1
gives a ReP(ˆO)Ph species.
The rhenium configurations,
a separate route described below. The H and C{ H} NMR
2
5
[26]
both corresponding to retention, were assigned by analogy to
many closely related reactions.
spectra showed patterns characteristic of h -C H X ligands.
5 4
The P{ H} spectrum showed three signals (Table 1), with the
C H PPh resonance not detectably coupled to the rhenium-
[
10, 12, 22, 24]
31
1
Standard methods
for assaying enantiomeric purities were not successful. How-
ever, complete retention is normally observed. The hydro-
genation enantioselectivities described below also require
5
4
2
bound phosphorus atoms. When this synthesis was conducted
3
1
1
with an excess of PPh Cl, a by-product formed. The P{ H}
2
3
1
1
very high enantiomer ratios. The P{ H} NMR spectra of all
NMR data suggested a species with a RePPh PPh linkage.
2
2
new complexes are summarized in Table 1. The PPh signals of
In a standard protocol for the synthesis of rhodium
3
complexes of chelating diphosphines,[ (R)-5, the rhodium
27]
norbornadiene complex [Rh(NBD)Cl]2,[
23b]
and AgPF were
1
Table 1. Summary of 31P{ H} NMR data.[a]
6
combined in THF at room temperature. As shown in
Scheme 2, workup gave the heterobimetallic rhenium/rhodi-
Complex
RePPh
3
Re(CH
2
)
n
PPh
2
X
C
±
5 4 2
H PPh X'
‡
� [b,g]
2
BF
4
13.3 (d)
� 5.5 (d)
5
um complex (S)-[(h -C H PPh )Re(NO)(PPh )(m-PPh )Rh-
5
4
2
3
2
2
2
J(P,P) ˆ 13 Hz
J(P,P) ˆ 13 Hz
� 48.3 (d)
‡
�
‡
�
(
NBD)] PF , [(S)-6 PF ], as a dark orange powder and
3[d,g]
6
6
19.5 (d)
±
±
2
2
THF hemisolvate in 92% yield. The structure followed
readily from the spectroscopic properties. Most diagnostic
J(P,P) ˆ 15 Hz
J(P,P) ˆ 15 Hz
4[d,g,i]
21.6 (d)
� 41.2 (d)
2
2
31
1
J(P,P) ˆ 16 Hz
J(P,P) ˆ 16 Hz
was the richly featured P{ H} NMR spectrum summarized in
5[d,g]
20.2 (d)
� 45.2 (d)
� 16.2 (s)
[18b]
Table 1 (and illustrated elsewhere).
Both diphenylphos-
2
2
J(P,P) ˆ 15 Hz
J(P,P) ˆ 15 Hz
1
‡
� [b,f,g]
phido signals exhibited large J(P,Rh) values (127, 183 Hz).
6
PF
6
9.8 (dd)
� 49.2 (ddd)
50.4 (ddd)
2
1
1
The C H PPh signal shifted downfield from that of precursor
J(P,P) ˆ 14 Hz,
J(P,Rh) ˆ 127 Hz,
J(P,Rh) ˆ 183 Hz,
5
4
2
3
2
2
J(P,P) ˆ 5 Hz
J(P,P) ˆ 19 Hz,
J(P,P) ˆ 19 Hz,
(R)-5 (d ˆ 50.4 vs � 16.2), and was coupled to both rhenium-
bound phosphorus atoms. All crystallization attempts gave
oils, but the crystal structure of the racemate is described
below.
2
3
J(P,P) ˆ 14 Hz
J(P,P) ˆ 5 Hz
1
8
4[b,g]
26.1 (s)
±
� 14.9 (s)
‡
� [b,h]
BF
4
21.7 (d)
30.2 (d)
±
3
3
J(P,P) ˆ 12 Hz
J(P,P) ˆ 12 Hz
9[e,h]
25.8 (d)
8.1 (d)
±
Six-membered chelates: We sought to compare a family of
catalysts. Hence, a second, homologous, series of complexes
was desired. One possibility was to retain the methyl carbon
that was removed by protonation in the first step of Scheme 2.
3
3
J(P,P) ˆ 8 Hz
J(P,P) ˆ 8 Hz
0[d,h,i]
1
1
28.3 (d)
10.0 (d)
±
2
2
J(P,P) ˆ 16 Hz
J(P,P) ˆ 16 Hz
1[c,h]
26.3 (d)
6.9 (dd)
� 17.7 (d)
3
3
3
J(P,P) ˆ 8 Hz
J(P,P) ˆ 3 Hz,
J(P,P) ˆ 3 Hz
‡
�
As shown in Scheme 3, either racemic or (S)-1 and Ph C BF
3
4
3
J(P,P) ˆ 8 Hz
‡
� [e,f,h]
were reacted in CH Cl at � 608C to generate the electrophilic
2
2
12
PF
6
20.2 (dd)
50.5 (ddd)
23.9 (ddd)
5
3
1
1
methylidene complexes, racemic or (S)-[(h -C H )Re(NO)-
5 5
J(P,P) ˆ 18 Hz,
J(P,Rh) ˆ 148 Hz,
J(P,Rh) ˆ 166 Hz,
3
2
2
‡
�
‡
�
[28]
J(P,P) ˆ 4 Hz
J(P,P) ˆ 34 Hz,
J(P,P) ˆ 34 Hz,
(PPh )(ˆCH )] BF , (7 BF ). Then PPh H was added.
3
2
4
4
2
3
3
J(P,P) ˆ 18 Hz
J(P,P) ˆ 4 Hz
Workups gave the new phosphonium salts, racemic or (S)-
5
‡
�
‡
�
4
[
a] At room temperature unless noted. [b] In CD
2
Cl
2
. [c] In C
6
D
6
. [d] In
[(h -C H )Re(NO)(PPh )(CH PPh H)] BF , (8 BF ), as
5 5 3 2 2 4
�
1
THF. [e] In CDCl
3
. [f] PF
6
� 144.0 (sep, J(P,F) ˆ 708 Hz). [g] 121 MHz.
orange to red prisms in 95 ± 98% yields. Similar syntheses of
related cationic species with ReCH PX linkages have been
[
h] 162 MHz. [i] At � 808C.
2
3
described.[ Deprotonations with tBuOK gave the trivalent
29]
5
phosphines, racemic or (S)-[(h -C H )Re(NO)(PPh )-
5
5
3
‡
�
(
R)-2 BF4 and (R)-3 were in normal ranges for this series of
(CH PPh )], (9), as orange to red needles in 91 ± 89% yields.
2 2
‡
�
compounds, and coupled to the other ligating phosphorus with
Racemic or (S)-8 BF4 could be stored as solids in air for
extended periods. Racemic or (S)-9 were much less air
sensitive than racemic or (R)-3, and solutions survived
exposures of several hours. Thus, the additional methylene
group attenuates the phosphorus lone pair reactivity. The
2
1
13
J(P,P) values of 13 ± 15 Hz. Other NMR ( H, C), as well as
IR, microanalytical, and polarimetric data are given in the
Experimental Section.
The cyclopentadienyl ligand of methyl complex 1 can be
lithiated and then alkylated.
similar sequence was investigated for elaborating (R)-3 to a
[
25]
31
1
‡
�
3
As shown in Scheme 2, a
P{ H} NMR spectra of 8 BF and 9 showed J(P,P) values
4
2
‡
�
that were 90 ± 60% of the J(P,P) values of 2 BF and 3
4
chelating diphosphine. Reaction with nBuLi (1.1 equiv,
(Table 1). The structures and configurations in Scheme 3
correspond to retention at rhenium. The addition of a carbon
788C) in THF gave a deep red solution, and a 31P{ H}
1
�
‡
�
NMR spectrum of an aliquot (Table 1) showed the clean
formation of a species that was assigned as the lithiocyclo-
nucleophile to 7 BF has been shown to proceed with
4
retention.[ The absolute configuration of (S)-9 is verified by
29b]
5
pentadienyl complex (R)-[(h -C H Li)Re(NO)(PPh )(PPh )],
a crystal structure below.
5
4
3
2
(
R)-4. It persisted, in separate experiments, for several days in
The conversion of racemic and (S)-9 to diphenylphosphi-
docyclopentadienyl complexes was attempted next. Reactions
solution at room temperature. Addition of PPh Cl (1.0 equiv)
2
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2017 $ 17.50+.50/0
2017

J. A. Gladysz et al.
FULL PAPER
+
BF4-
a)
b)
Re
Re
ON
PPh3
ON
PPh3
CH3
CH PPh H
2
2
(S)-1 or
(S)-8 or
racemate
racemate
c)
Li
d)
Re
Re
ON
PPh3
ON
PPh3
CH2PPh2
CH2PPh2
(S)-10 or
(S)-9 or
racemate
racemate
e)
+
Ph2P
Ph2P
Rh
f)
PF6-
Re
Re
ON
PPh3
NO
Ph2P
C
H2
CH2PPh2
PPh3
Figure 1. 31P{ H} NMR spectra of 11 (left) and 12 PF
1
‡
�
6
(right).
(S)-12+ PF6-
(S)-11 or
racemate
or racemate
Scheme 3. Syntheses of the racemic and non-racemic six-membered
Racemic five-membered chelates: The synthesis of (S)-
‡
�
chelating diphosphines. a) Ph
to 258C; c) tBuOK, THF, 258C; d) tBuLi, THF, � 60 to � 108C; e) PPh
THF, � 78 to 258C; f) [Rh(NBD)Cl] , AgPF , THF, 208C.
3
C BF
4
, CH
2
Cl
2
, � 608C; b) PPh
2
H, � 60
‡
�
6
PF6 in Scheme 2 represents a second-generation ap-
2
Cl,
proach. The first generation approach, shown in Scheme 4,
2
6
5
was initiated at a time when syntheses of adducts of [(h -
n‡
C H )Re(NO)(PPh )] and phosphorus-donor ligands were
5
5
3
still being optimized.[ It is described here because of certain
18b]
with tBuLi (1.2 equiv, ꢀ � 308C, then warming) in THF gave
3
1
1
deep red solutions. The P{ H} NMR spectra of aliquots
showed the clean formation of new signals (Table 1) that were
attributed to the lithiocyclopentadienyl complexes, racemic
Ph2P
5
and (R)-[(h -C H Li)Re(NO)(PPh )(CH PPh )] (10, Sche-
a)
5
4
3
2
2
Re
Re
H
me 3). Additions of PPh Cl (1.0 equiv) and workups gave the
2
PPh3
PPh3
ON
ON
5
target complexes, racemic and (S)-[(h -C H PPh )Re(NO)-
5
4
2
OTs
(
PPh )(CH PPh )], (11), as orange-red solids in 70 ± 68%
3 2 2
(S)-14 or
(R)-13 or
yields. These could be exposed to air for brief periods, but
racemate
racemate
were much more sensitive than the precursors, racemic and
S)-9. The 31P{ H} NMR spectrum of 11 (Table 1) was better
1
(
b)
resolved than that of lower homologue 5, and only the RePPh3
and C H PPh signals were not detectably coupled.
Syntheses of rhodium chelate complexes were investigated.
As shown in Scheme 3, racemic and (S)-11 were treated with
5
4
2
Ph2P
Ph2P
c)
Re
Re
_
ON
PPh3
ON
PPh3
[
Rh(NBD)Cl] and AgPF . Workups gave the heterobimetal-
Li+
2
6
PPh2
5
lic rhenium/rhodium complexes, racemic and (S)-[(h -
5
15
‡
�
C H PPh )Re(NO)(PPh )(m-CH PPh )Rh(NBD)] PF ,
5
4
2
�
3
2
2
6
racemate
‡
(
12 PF ), as brown or brownish red solids in 95 ± 82% yields.
6
These showed good air and thermal stabilities. The structures
followed readily from the spectroscopic properties, the most
diagnostic of which were the P{ H} NMR data (Table 1).
Figure 1 compares the highly informative coupling patterns
Ph2P
H
3
1
1
_
Re
ON
PPh3
with those of precursor 11. Both PPh Rh signals exhibit large
J(P,Rh) values, and are markedly downfield from their
OTs
2
1
F
counterparts in 11. Deep red prisms of a CH Cl monosolvate
of the racemate were obtained, and the crystal structure is
described below.
2
2
Scheme 4. Synthesis of the racemic five-membered chelating diphosphine
[19]
5
.
a) LiPPh
2
, THF, 208C, partial retention; b) nBuLi, THF, � 158C;
c) PPh
2
Cl, THF, � 788C.
2018
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2018 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

ªChiral-at-Rheniumº Complexes
2015±2027
5
novel features, and to accurately represent how some key data
were obtained.
deep red solution that was believed to contain [(h -
�
‡
C H PPh )Re(NO)(PPh )] Li (15). Addition of PPh Cl
5
4
2
3
2
5
The tosylate complexes, racemic and (R)-[(h -C H )Re-
gave the diphenylphosphido complex 5 in 68% yield
after workup. A sequence starting with the (R)-14 of un-
known enantiomeric purity also gave racemic 5. This is
in accord with earlier precedent,[ and shows that the
diphenylphosphido moiety does not impart any special
configurational stability to 15. Racemic 5 was converted
to racemic 6 PF6 as described for the non-racemic analogues
in Scheme 2. Workup gave a THF monosolvate in 83%
yield. Dark red prisms of a CHCl3 disolvate were also
obtained.
5
5
[
19, 30]
(
NO)(PPh )(OTs)], (13),
were treated with LiPPh in
2
3
THF. The formation of diphenylphosphido complexes, race-
31]
mic and (R)-3, was anticipated. Workups gave homogeneous
1
13
products in high yields. However, the H and C NMR
5
spectra showed patterns characteristic of h -C H X li-
5
4
[
25, 26]
31
1
‡
�
gands.
The P{ H} NMR spectra (Table 1) exhibited
two uncoupled signals, with chemical shifts close to those of
the PPh and C H PPh ligands in 5 and 11. The IR and
3
5
4
2
1
H NMR spectra showed plausible signals for hydride ligands
�
1
(
nÄ ReH ˆ 1950cm ; d ˆ � 9.56) that closely matched those of
5
[
3
1
]
the parent complex [(h -C H )Re(NO)(PPh )(H)].
cordingly, the products were assigned to the structure [(h -
C H PPh )Re(NO)(PPh )(H)] (14; 85 ± 79%). The triflate
Ac-
5
5
3
5
Catalyst structure: The crystal structures of (S)-9 ´ C H and
the racemic rhenium-rhodium chelates 6 PF ´ (CHCl ) and
12 PF6 ´ CH Cl were determined as outlined in Table 2 and
6
6
‡
�
5
4
2
3
6
3 2
5
[23c, 30]
‡
�
complex [(h -C H )Re(NO)(PPh )(OTf)]
reacted under
5
5
3
2 2
similar conditions to give a mixture of 3 and 14.
the Experimental Section. The first is shown in Figure 2
(bottom), together with that previously found for the diphe-
nylphosphido complex 3 (top).[ The structure confirms the
absolute configuration, and the overall stereochemistry
(retention) for the first two steps in Scheme 3. It also exhibits
Thus, the first step of Scheme 4 unexpectedly gave a
diphenylphosphidocyclopentadienyl ligand that was, howev-
er, ultimately desired. Similar nucleophile-initiated sequences
that lead to substituted cyclopentadienyl ligands and displace-
ment of coordinated ligands by cyclopentadienyl-derived
13a]
a LRe�CH X conformation similar to those of several related
2
[
32, 33]
[15, 29b, 34]
hydride have been reported.
From the limited data
complexes.
As illustrated by Newman-type projection
available, we presume that addition to the cyclopentadienyl
ligand precedes hydride migra-
H in Figure 3 (and quantified by the P1/N1-Re-C1-P2 torsion
tion, as shown in structure F in
Scheme 4. However, there is no
Table 2. Crystallographic data.
precedent for the stereochem-
istry at the metal. The sample
‡
�
‡
�
6
(
S)-9 ´ C
6
H
6
6
PF
6
´ (CHCl
3
)
2
12 PF
´ CH
2 2
Cl
molecular formula
molecular weight
T[8C]
diffractometer
radiation [Š]
crystal system
space group
unit cell dimensions:
a [Š]
C
820.87
42
H
38NOP
2
Re
C
56
H
49Cl
6
F
6
NOP
4
ReRh
C
56
H
51Cl
2
F
6
NOP ReRh
4
derived from (R)-13 kept some
1461.65
15
Syntex P1-bar
MoKa
monoclinic
P2
1387.32
2
2
optical activity ([a] ˆ 898).
5
89
� 100(2)
� 100(2)
The ORD spectrum was posi-
tive at 650 ± 375 nm and nega-
tive at 375 ± 250 nm, and the
CD spectrum was positive at
Nonius MACH3
MoKa
Nonius MACH3
MoKa
monoclinic
P2
monoclinic
P2 /c
1
1
1
/c
6
50 ± 450 nm and 425 ± 275 nm
and negative at 450 ±
25 nm.[18b, 19] This suggested
dominant retention of confi-
12.086(2)
8.559(2)
17.785(4)
105.87(3)
1769.6(6)
2
12.980(4)
15.832(4)
29.117(7)
91.70(3)
5980.9(2)
4
13.392(3)
11.155(2)
34.445(7)
98.46(3)
5089.7(18)
4
b [Š]
c [Š]
4
b [8]
V [Š ]
3
[
29b, 30]
guration [(S)-14].
Howev-
Z
er, the magnitude of the rota-
tion is low for chiral rhenium
compounds, and the enantio-
meric purity could not be as-
sayed.
� 3
1
c
[gcm
]
1.54
3.56
1.66
4.40
1.81
3.05
m [mm 1]
�
data collection method
crystal dimensions [mm]
V range [8]
V scans
0.2 Â 0.2 Â 0.2
ꢀ 25
V scans
0.3 Â 0.3 Â 0.3
ꢀ 25
V scans
0.4 Â 0.4 Â 0.3
ꢀ 25
range/indices (h,k,l)
� 14,14;
� 10,10;
� 21,21
5988
0,15;
0,16;
0,27
10957
� 15,15;
� 13,13;
� 40,40
18892
The elaboration of the hy-
dride ligand in 14 was attempt-
ed. The addition of nBuLi to
reflections measured
unique reflections
reflections observed
refined parameters
refinement
5290
10957
7201 [I > 3s(I)]
1269
8951
5
the parent complex [(h -C H )-
5
5
4715 [I > 2s(I)]
424
least-squares on F2
0.0392
R
wR
6405 [I > 2s(I)]
649
least-squares on F2
Re(NO)(PPh )(H)]
(THF,
3
�
158C) generates the red, rhe-
least-squares on F
5
Rint
R indices
±
0.0840
nium-centered anion [(h -
�
‡ [31]
1
ˆ 0.0355
R ˆ 0.0740
R
1
ˆ 0.0404
C H )Re(NO)(PPh )] Li .
5
5
3
[
I > 2s(I)]
2
ˆ 0.0842
[R
w
] ˆ 0.0830
wR
2
ˆ 0.0807
This in turn reacts with a variety
of electrophiles to give new
R indices (all data)
R
1
ˆ 0.0448
±
R
1
ˆ 0.0801
wR ˆ 0.0934
2
wR ˆ 0.1036
2
s
complexes. An analogous
reaction of racemic 14 and
nBuLi (Scheme 4) gave
goodness of fit
largest diff. peak, hole [eŠ 3]
1.003
±
1.041
�
5.38[a]/ ±
2.184/ � 1.653
1.838/ � 1.632
a
[a] 1.03 Š from Re.
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2019 $ 17.50+.50/0
2019

J. A. Gladysz et al.
FULL PAPER
Table 3. Key distances [Š] and angles [8] in 3 and (S)-9 ´ C
6 6
H .
3
(S)-9 ´ C H
6 6
Re�N1
1.738(10)
2.358(3)
2.461(3)
±
1.773(7)
2.352(2)
±
2.170(8)
1.845(8)
1.949
2.324(8)
92.5(2)
174.5(6)
84.0(4)
±
Re�P1
Re�P2
Re�C1
C1�P2
±
Re�Cp(centroid)
Re�C13
1.944
2.287(16)
91.5(4)
177.9(10)
N1-Re-P1
Re-N1-O1
C1-Re-C13
P2-Re-C13
N1-Re-C1
±
91.4(8)
±
92.5(4)
97.2(3)
±
N1-Re-P2
P1-Re-C1
P2-Re-P1
±
87.6(2)
±
92.5(1)
Re-C1-P2
-
±
±
±
±
112.1(4)
P1-Re-C1-P2
N1-Re-C1-P2
Re-C1-P2-LP
Re-C1-P2-C50
Re-C1-P2-C60
P1-Re-P2-LP
N1-Re-P2-LP
P1-Re-P2-C50
P1-Re-P2-C60
N1-Re-P2-C50
N1-Re-P2-C60
� 159.9
� 67.6
� 49.1
� 176.9
±
78.8
±
60.7
30.9
62.6
175.9
154.3
±
±
±
±
� 92.5
±
6 6
Figure 2. Molecular structure of 3 (top) and (S)-9 C H (bottom).
‡
�
The structures of the cations of chelates 6 PF and
6
‡
�
1
2 PF6 are shown in Figure 4. Selected distances and
angles are compared in Table 4. With one exception (below),
the bond lengths or angles common to both chelate rings are
similar. For example, both PRhP angles are close to that
expected for an idealized square planar rhodium geometry
(
91.35(12), 95.41(7) Š). However, the rhodium ± rhenium
distance is 10% greater in the larger six-membered ring
4.505 vs 4.068 Š). Protonation or alkylation of the phos-
phorus lone pair in 3 would relieve repulsive interactions and
(
[13b]
give a much shorter rhenium ± phosphorus bond.
ever, the introduction of rhodium in 6 PF6 leads to a
slightly longer rhenium ± phosphorus bond (2.487(3) vs
How-
‡
�
Figure 3. Selected conformational features in crystal structures.
2
.461(3) Š).
‡
�
‡
�
6
Newman-type projections of 6 PF6 and 12 PF are
angles in Table 3), the larger X group occupies the interstice
between the cyclopentadienyl and nitrosyl ligands. Numerous
NMR and computational data have shown this region to be
illustrated in Figure 3 (I, J). Due to the geometrical constraint
‡
�
of the chelate, the LRePPh conformations of 3 and 6 PF
2
6
(G, I) differ significantly. However, since rhodium is now the
largest group on phosphorus, that in I would be favored even
[
35, 36]
the least congested.
In the crystal structure of 3 (Figure 2, top), the smallest
group on the diphenylphosphido ligand, the lone pair, is
directed into the interstice between the nitrosyl and triphe-
nylphosphine ligands (see G, Figure 3). This is known to be
in the absence of a chelate. In contrast, the LReCH P
conformations in (S)-9 and 12 PF6 (H, J) must change only
2
‡
�
slightly, as reflected by the torsion angles (L ˆ Ph P/ON:
3
� 159.98/ � 67.68 vs � 157.38/ � 68.38). The situation with
the adjacent ReCH �PPh linkage is similar. Geometrical
[
35]
the most congested region. However, there is an additional
electronic driving force. In this conformation, the torsion
angle between the lone pair and rhenium fragment HOMO
shown in C (Scheme 1) is about 608. The high degree of
2
2
constraints require the rhodium in J to assume the phos-
phorus lone pair position in H (see also Figure 2, bottom).[
Accordingly, the ReC ± PPh torsion angles in (S)-9 and
36]
[
13a]
‡
�
orthogonality lessens repulsive interactions.
Key metrical
12 PF6 are very close (78.88/ � 176.98 vs 73.18/179.48 or
parameters of 3 and (S)-9 are compared in Table 3. The effect
of rhodium chelate formation upon these core structures is of
obvious interest.
� 180.68). However, the ReCH ± PRh conformation is
2
opposite to what would be expected in the absence of a
chelate.[
36b]
2020
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2020 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

ªChiral-at-Rheniumº Complexes
2015±2027
‡
�
Table 4. Key distances [Š] and angles [8] in 6 PF ´ (CHCl ) and
6 3 2
‡
�
1
2 PF ´ CH Cl .
6 2 2
‡
�
‡
�
6
6
PF
6
´ (CHCl
3
)
2
2 2
12 PF ´ CH Cl
Re�N1
Re�P1
Re�P2
Re�C1
C1�P2
1.731(13)
2.392(3)
2.487(3)
±
1.760(7)
2.365(2)
±
2.183(7)
1.825(7)
1.939
2.253(7)
1.798(7)
2.295(2)
2.360(2)
4.505
89.1(2)
100.5(3)
±
89.8(2)
±
117.4(3)
±
99.4(3)
±
±
123.3(3)
113.9(3)
95.41(7)
113.3(2)
108.5(2)
111.0(2)
115.1(2)
� 61.8
� 170.9
� 157.3
� 68.3
±
±
Re�Cp(centroid)
Re�C24
1.925
2.217(12)
1.805(12)
2.276(4)
2.380(3)
4.068
90.8(3)
±
99.8(4)
±
99.55(12)
±
113.4(3)
±
87.4(3)
121.9(4)
117.1(4)
110.6(4)
91.35(12)
102.4(4)
108.7(4)
113.3(5)
114.8(4)
P3�C24
Rh�P3
Rh�P2
Re�Rh
N1-Re-P1
N1-Re-C1
N1-Re-P2
C1-Re-P1
P2-Re-P1
Re-C1-P2
Re-P2-Rh
C24-Re-C1
C24-Re-P2
Rh-P2-C1
Re-C24-P3
Rh-P3-C24
P2-Rh-P3
Rh-P2-C50
Rh-P2-C60
Rh-P3-C70
Rh-P3-C80
Re-C24-P3-C70
Re-C24-P3-C80
P1-Re-C1-P2
N1-Re-C1-P2
N1-Re-P2-C50
N1-Re-P2-C60
P1-Re-P2-C50
P1-Re-P2-C60
P1-Re-P2-Rh
Re-C1-P2-Rh
Re-C1-P2-C50
Re-C1-P2-C60
� 69.4
179.5
±
‡
�
Figure 4. Structures of the cations of
6 PF ´ (CHCl ) (top) and
6 3 2
±
‡
�
12 PF ´ CH Cl (bottom).
6 2 2
17.6
136.7
� 74.9
44.2
167.2
±
±
±
±
±
Figure 4 illustrates the overall conformation of the chelate
‡
�
‡
�
rings in 6 PF6 and 12 PF . A close inspection reveals an
6
58.1
73.1
179.4
uncanny resemblance to the classical envelope and chair
conformations of cyclopentane and cyclohexane, respectively.
The similarities are highlighted in K and L in Figure 3. The
latter features a striking array of three syn-axial phenyl and
nitrosyl groups. We speculate that the essentially unrestricted
rotation about the rhenium ± cyclopentadienyl axis allows the
Re(CH ) PRhP linkage to attain conformational energy
±
±
‡
�
quadrants in Figure 4). The phenyl rings in 6 PF define a
6
3
P M pattern, which is found in only 10% of the five
membered-chelates with envelope conformations of the same
2
n
minima similar to those of the corresponding carbocycles.
Accordingly, the C24-Re-P2 and C24-Re-C1 angles, which
reflect this degree of freedom, are quite different in the two
chelates (87.4(3)8 vs 99.4(3)8).
configuration (most are M4 or M P).
3
[5a]
Furthermore, the
phenyl ring containing C50 adopts a normally forbidden
orientation, apparently to minimize interactions with the
pseudoaxial nitrosyl group. In contrast, the phenyl rings in
‡
�
‡
�
Views of 6 PF6 and 12 PF without the norbornadiene
6
‡
�
2
2
1
2 PF6 adopt a P M configuration. Derivations of these
ligand are collected in Figure 5. An overlay (top) in which the
chiral rhenium moieties are superimposed as closely as
possible highlights the different chiral environments at
rhodium. The chiral environments are then shown from the
perspective of the ligands on rhodium (middle and bottom
structures). Brunner has carefully analyzed such chiral pock-
ets in complexes of chiral chelating bis(diphenylphosphine-
relationships are given as Supporting Information (Table 1-S).
The most important point is that they validate our hypothesis
that chelates of the type B (Scheme 1) should provide unique
and heretofore inaccessible types of steric environments for
enantioselective catalysts.
[
5a]
s). He has classified the phenyl ring orientations into four
types based upon torsion angle relationships: face ± P, edge ±
P, edge ± M, and face ± M, where P and M are helical chirality
descriptors.
Catalytic enantioselective hydrogenations: Many chiral cati-
onic rhodium diphosphine complexes have been evaluated as
catalyst precursors for enantioselective hydrogenations of
Some are very effective with a,b-unsatu-
rated carboxylic acids and esters that bear a-acetamido
alkenes.[
1, 4a,c±e, 16, 27a]
‡
�
‡
�
Complexes 6 PF and 12 PF both feature two phenyl
6
6
rings with face-orientations with respect to rhodium (upper
substituents. These afford products of obvious interest,
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2021 $ 17.50+.50/0
2021

J. A. Gladysz et al.
FULL PAPER
Table 5. Catalytic enantioselective hydrogenation of dehydroamino
[
a]
acids.
Entry
Educt
Catalyst
T
Pressure
[bar]
Yield
[%]
ee
[%]
[
8C]
‡
�
�
�
�
1
2
3
4
5
6
7
8
9
0
1
2
4
5
16a
16b
16c
16d
16a
16b
16c
16d
16b
16b
16b
16b
16b
16b
(S)-6 PF
6
6
6
6
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
20 ± 23
1
1
1
1
1
1
1
1
0.3
10
30
90
1
82
70
94
86
93
86
98
96
±
±
±
±
75
82
92
93
88
82
62
72
65
60
76
60
46
40
40
72
‡
(S)-6 PF
‡
(S)-6 PF
‡
(S)-6 PF
‡
�
�
�
�
�
�
�
�
�
�
(S)-12 PF
(S)-12 PF
(S)-12 PF
(S)-12 PF
(S)-12 PF
(S)-12 PF
(S)-12 PF
(S)-12 PF
6
6
6
6
6
6
6
6
6
6
‡
‡
‡
‡
‡
1
1
1
1
1
‡
‡
‡
(S)-12 PF
(S)-12 PF
� 5
‡
45
1
[a] Conditions and analytical methods are detailed in the Experimental
Section.
This follows intuitively from the closer proximity of the chiral
rhenium to one diphenylphosphido group and in turn the
rhodium.
The effect of hydrogen pressure upon enantioselectivity
‡
�
was examined with 16b and (S)-12 PF6 (entries 9, 6, 10, 11,
2). An inverse dependence was observed, in other words
lower ee values of 17b at higher pressures. This has been
1
‡
�
Figure 5. Top: Superposition of the cations of 6 PF
6
´ (CHCl
3
)
2
and
‡
�
12
PF
6
´ CH
2
Cl
2
, omitting the norbornadiene and PPh
Middle and bottom: View of the chiral pockets of 6 and 12 from the
perspective of ligands on rhodium, omitting the norbornadiene and one
3
phenyl rings.
found for many other rhodium catalysts with chiral chelating
‡
‡
diphosphines.[
16d, e]
The ee values also decreased at lower
temperature (entries 6, 14, 15). The mechanistic basis for
3
PPh phenyl ring.
these phenomena has been analyzed in detail.[
16c±e]
They
strongly suggest that our new catalysts exhibit analogous CˆC
enantioface binding equilibria and reactivity patterns. Turn-
over frequencies were measured by monitoring hydrogen
uptake with a gas burette. Values ranged from 0.028 to
protected a-amino acids. As summarized in entries 1 ± 8
of Table 5, four such substrates (16a ± d) were treated with
1
atm of hydrogen in the presence of 0.5 mol% of the
‡
� [19]
‡
�
� 1
� 1
‡
�
� 1
rhenium ± rhodium chelates (S)-6 PF
and (S)-12 PF6
0.083 s (100 ± 300 h ) for (S)-6 PF
6
, and 0.43 ± 0.83 s
�
. Thus, the less selective
6
6
�
1
‡
in THF at ambient temperature. Workups gave the amino
acid derivatives 17a ± d in 70 ± 98% yields. All spectroscopic
and chromatographic probes indicated quantitative reac-
tions.
(1550 ± 2970 h ) for (S)-12 PF
catalyst is more reactive. Complex (S)-6 PF
‡
�
gave similar
Cl and methanol, but
2
6
yields and turnover frequencies in CH
enantioselectivities were not assayed.
2
The absolute configurations of 17a ± d were assigned
polarimetrically. The enantiomeric purities were determined
by chiral chromatographic methods as described in the
The yields and catalyst loadings (0.5 mol%) in Table 5
translate to turnover numbers of 140 ± 195. We sought to
probe the efficacy of lower loadings. Accordingly, the reaction
in entry 6 was repeated in a more concentrated solution with
0.05 mol% catalyst. Hydrogen was taken up over the course
‡
�
Experimental Section. The samples derived from (S)-6 PF6
entries 1 ± 4) were first analyzed by GLC, and then stored for
fourteen years and analyzed by GLC or HPLC. Compound
7a in entry 1, originally reported as 98% ee from a non-
(
�
1
� 1
of 1.5 h (TOF 0.50 s or 1800 h ). Workup gave 17b in 81%
yield (70% ee), corresponding to a turnover number of 1620.
However, considering the spectroscopically quantitative na-
ture of these reactions, we believe that the true value is closer
1
[
18]
baseline enantiomer separation, was found to be 92% ee.
The other values were in agreement. In all cases, the five-
‡
�
‡
�
‡
�
membered chelate, (S)-6 PF , gave higher ee values than the
to 2000. Regardless, (S)-6 PF
6
and (S)-12 PF
6
may be used
6
‡
�
six-membered chelate, (S)-12 PF (entries 1 ± 4 vs 5 ± 8).
at very low loadings.
6
2022
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2022 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

ªChiral-at-Rheniumº Complexes
2015±2027
centers are easily modified.[39] The corresponding penta-
methylcyclopentadienyl complexes are also available in
enantiomerically pure form,[ and tetra- or trisubstituted
homologues should be similarly accessible. It would be easy to
replace the methylene group in the larger chelating ligand 11
by either configuration of a CHR stereocenter.[ There are
also many ways by which 1 could be elaborated on a solid
support, or that mixed phosphorus/nitrogen or phosphorus/
sulfur donors could be accessed.
Discussion
40]
Schemes 2 and 3 document the ready availability of a novel
new class of chiral chelating diphosphines [(R)-5, (S)-11],
which, due to the presence of a chiral transition metal in the
backbone, present heretofore inaccessible types of steric and
electronic environments. These give rhodium adducts that are
enantioselective hydrogenation catalyst precursors and show
very promising first-generation effectiveness, approaching
some the best performance benchmarks in the literature.
There is no reason to doubt that optimization of this catalyst
class would not match or exceed these benchmarks. There-
fore, we focus the first part of this discussion on comparisons
with ferrocene-based chiral chelating diphosphines (A, A';
Scheme 1). A logical starting question is ªhow easily can each
type of chelating diphosphine be synthesized?º
29c]
Some related heterobimetallic complexes that have been
prepared by other groups deserve emphasis. Brunner has
reported the chiral molybdenum ± rhodium complex 21, which
is shown in Scheme 6 and features molybdenum, carbon, and
nitrogen stereocenters. It catalyzed the hydrosilylation of
ketones, but only modest ee values were obtained.[ Moïse
has synthesized novel chiral tantalum complexes with diphe-
nylphosphido and diphenylphosphidocyclopentadienyl li-
gands.[ He has shown that these can chelate to other metals,
as illustrated by 22 in Scheme 6. However, only racemic
8a]
The conversion of commercially available Re (CO) to the
non-racemic methyl complex (S)-1 and then diphosphines
R)-5 or (S)-11 requires eleven steps (including two tandem
2
10
41]
(
steps, such as the BuLi/PPh Cl sequences). The overall yields
2
are 30% and 32%, respectively. Scheme 5 summarizes
analogous statistics for three representative ferrocene-based
[
4b, 37b]
phosphines, 18 ± 20.
These require seven-fifteen steps
from ferrocene, with 15 ± 40% overall yields. Thus, our
chelates compare favorably from a preparative standpoint.
Our starting material is more expensive than ferrocene, but
this becomes less of a factor over a multistep sequence.
Furthermore, both enantiomeric series of rhenium complexes
[
19]
Scheme 6. Other relevant literature compounds.
are equally available.
Another important comparison involves hydrogenation
enantioselectivities. The best of the dozens of rhodium
catalysts derived from ligand types A/A' deliver 17a ± c in
complexes have been described to date. Finally, the mixed
phosphorus/nitrogen donor ligand 23 has been reported by
[
37b, 38]
[42]
9
6
7 ± 99% ee,
as compared to 88 ± 93% ee with (S)-
Helmchen. It features a cyclopentadienyl Mn(CO) moiety
3
‡
�
PF . Of course, other types of chiral chelating diphos-
with planar chirality, and other stereogenic elements. Many
would have feared that the manganese would be too labile for
most types of catalytic reactions. Nonetheless, this ligand is
extremely effective for palladium-catalyzed enantioselective
allylic substitutions.
In conclusion, this study has established the ready avail-
ability and efficacy of a new and architecturally novel type of
chiral chelating diphosphines for metal-catalyzed enantiose-
lective organic reactions. The general strategy, exemplified by
B in Scheme 1, can easily be extended to a broad new family
of chelates. By every criterion, these appear capable of
impacting catalysis analogously to ferrocene-based chiral
chelates of the types A and A'. Additional types of catalyst
precursors and applications will
6
phines give similar or still higher values. From this extensive
literature, we emphasize the DuPHOS ligand family.[
Here, care was also taken to document TON values, which
for preparative reactions (>10 g) were routinely 10000. This
is larger than the maximum we have demonstrated (1620), but
we are confident our catalysts would match it on similar time
and mass scales. Thus, although an about 10% improvement
in ee values is needed to render our catalyst family compet-
itive, there are no identifiable drawbacks or disadvantages
associated with the rhenium in the chelate backbone.
16a,b]
In addition, the rhenium-containing chelating diphosphines
have many intrinsic diversity elements. All three phosphorus
be reported in the near fu-
ture.
R Ph2P
Ar2P
[17]
NMe2
Ph2P
Ph2P
Ph
PPh2
PPh2
Re
Re
Fe
Fe
PAr2
Fe
Fe
ON
PPh3 ON
PPh3
Experimental Section
PPh2
CH2PPh2
Ph
PPh2
General data: All reactions except
hydrogenations were carried out un-
der dry N atmospheres with glassware
2
that had been oven-dried (1008C),
assembled while warm, and cooled
under vacuum. New compounds were
characterized as follows: IR and NMR
NMe2
19
7 steps from
(
R)-5
(S)-11
18
20
10 steps from
Re2(CO)10,
10 steps from
Re2(CO)10,
32% overall
7-8 steps from
ferrocene,
15-35% overall
15 steps from
ferrocene,
25-35% overall
[37a]
[
37b]
[4b]
ferrocene,
40% overall
30% overall
Scheme 5. Comparison of syntheses of representative ªmetal-containingº chelating diphosphines.
Chem. Eur. J. 2001, 7, No. 9 ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001 0947-6539/01/0709-2023 $ 17.50+.50/0
2023

J. A. Gladysz et al.
FULL PAPER
� 1
spectra, standard FT instruments; optical rotations, Perkin ± Elmer 241
spectropolarimeter; ORD/CD spectra, JASCO J-20C spectrophotometer;
mass spectra, VG 770 or Micromass Zabspec instruments; microanalyses,
Schwarzkopf Laboratories or in-house service (Carlo Erba EA 1110).
Chromatography was conducted with standard instruments under condi-
tions detailed below.
those of the racemate. [a] 52 28 9 ˆ � 898 (c ˆ 1.0 mg mL , THF). See text for
comments on the enantiomeric purity and configurational assignment.
5
[
(h -C
5
H
4
PPh
2
)Re(NO)(PPh
3
)(PPh
2
)] (5)
A) A Schlenk tube was charged with 14 (0.100 g, 0.140 mmol)[13a] and THF
(
10 mL), capped with a septum, and cooled to � 158C (ethylene glycol/
CO ). Then nBuLi (1.4m in hexane; 0.011 mL, 0.15 mmol) was added with
stirring. The light yellow solution turned deep red. After 20 min, the tube
was transferred to a � 788C bath (acetone/N ). Then PPh Cl (0.027 mL,
.033 g, 0.15 mmol) was added dropwise with stirring, giving an orange
solution. The cold bath was allowed to warm. After 4 ± 6 h, the volatiles
were removed by oil-pump vacuum. The residue was extracted with
benzene. The extract was filtered through a Celite plug. The filtrate was
concentrated, and hexanes were added by vapor diffusion. Red flower-like
crystals slowly formed, which were collected by filtration and dried
Solvents were treated as follows and stored under N
distilled from Sicapent (Fluka); benzene, distilled from CaH
ether, and toluene, distilled from Na/OˆCPh
; pentane and hexane,
distilled from Na; methanol and ethanol, distilled from Mg; chlorobenzene
Fluka, >99.5%), stored over molecular sieves; CD Cl and CDCl , trap-
to-trap distilled from molecular sieves; C and D [THF], trap-to-trap
distilled from Na/Pb alloy. Reagents were treated as follows:
Rh(NBD)Cl] (Strem, >99%), used as received; PPh H (>95%), AgPF
(Fluka), used as received; PPh Cl (Fluka,
2
: CH
2
Cl
2 3
and CHCl ,
2
2
; THF, diethyl
2
2
2
0
(
2
2
3
6
D
6
8
[
(
2
2
6
‡
�
>99%), and Ph
3
C BF
4
2
�
3
(
10 Torr, 568C, 12 h) to give 5 ´ (C
196 ± 1978C decomp; elemental analysis calcd (%) for: C47
0.5 (952.0): C 63.08, H 4.45, P 9.76; found: C 63.52, H 4.58, P 9.02; IR
6
H
6
)
0.5 (0.089 g, 0.093 mmol, 68%). M.p.
ꢁ
97%), freshly vacuum distilled; nBuLi (Fluka, ꢁ1.6 m in hexanes) and
[43]
H
39NOP Re ´
3
tBuLi (Fluka, ꢁ1.5 m in pentane), standardized;
hydrogenation sub-
6 6
(C H )
strates, used as received (16a,b; Fluka) or prepared by standard procedures
�
1
1
16c,d);[ others, used as received from common commercial sources.
44]
(KBr, cm ): nÄ ˆ 1656 (s, NO); H NMR (300 MHz, [D ]THF, 288C, TMS):
6 6
), 7.26 (brs, 0.5C H ), 5.34, 5.12, 4.59,
8
(
d ˆ 7.47 ± 7.27, 7.16 ± 6.92 (2m, 7C
6
H
5
5
‡
�
‡
�
]:[19] A Schlenk
(
S)-[(h -C
5
H
5
)Re(NO)(PPh
3
)(PPh
2
H)] BF
4
[(S)-2 BF
4
13
1
3
.02 (4brs, C
5
H
4
); C{ H} NMR (75 MHz, [D
8
]THF, 288C, TMS): PPh
3
and
5
[9]
flask was charged with (R)-[(h -C
5
H
5
)Re(NO)(PPh
3
)(CH
3
)], [(R)-1],
2
2
PPh
1
2
at d ˆ 139.2 (d, J(C,P) ˆ 11 Hz, o), 136.7 (d, J(C,P) ˆ 11 Hz, o'),
(
(
0.833 g, 1.49 mmol) and chlorobenzene (150 mL), and cooled to � 418C
CH CN/CO ). Then HBF (5.5m in diethyl ether; 0.271 mL, 1.5 mmol) was
H (0.416 g, 2.24 mmol) was added to
1
36.1 (d, J(C,P) ˆ 54 Hz, i), 135.2 ± 133.9, 130.8 ± 126.0 (2brm, Ph); 129.0
1
3
2
4
(
s, C
H
); C
5
H
at 101.4 (d, J(C,P) ˆ 18 Hz, CP), 93.8 (d, J(C,P) ˆ 2 Hz),
6
6
4
added with stirring. After 10 min, PPh
2
31
1
9
2
�
1.2 (d, J(C,P) ˆ 3 Hz), 91.2 (s), 90.9 (s); P{ H} NMR (121 MHz, [D
8
]THF,
PPh ),
the dark red solution. The cold bath was allowed to warm. After 12 h, the
mixture was slowly added to diethyl ether (400 mL). The tan powder was
isolated by filtration, washed with pentane (300 mL) and dried by oil pump
2
88C, H
3
PO
4
): d ˆ 20.2 (d, J(P,P) ˆ 15 Hz, PPh
3
), � 16.2 (s, C
5
H
4
2
2
45.2 (d, J(P,P) ˆ 15 Hz, RePPh
2
).
B)[19] A Schlenk tube was charged with (S)-3 (0.500 g, 0.620 mmol) and
THF (30 mL), capped with a septum, and cooled to � 788C. Then nBuLi
(1.6m in hexanes, 0.43 mL, 0.69 mmol) was slowly added with stirring. After
P NMR (Table 1; complete formation
of (S)-4). Then PPh Cl was added (0.110 mL, 0.140 g, 0.620 mmol; caution:
‡
�
4
(1.078 g, 1.320 mmol, 89%). M.p. 210 ± 2158C
vacuum to give (S)-2 BF
decomp; elemental analysis calcd (%) for C35
H
31BF
4
NOP
2
Re (816.6) for:
2
2
� 1
C 51.48, H 3.83; found: C 51.59, H 3.96; [a]589 ˆ � 878 (c ˆ 0.11 mgmL
,
15 min, an aliquot was assayed by 31
); 1H NMR (300 MHz, CD
CH
2
Cl
2
2
3
Cl
2
, 258C, TMS): d ˆ 7.60 ± 7.04 (m,
1
5
C
6
H
5
), 7.34 (dd, J(H,P) ˆ 392 Hz, J(H,P) ˆ 5.4 Hz, PH), 5.33 (s, C
5
H
5
);
2
3
1
1
2
any excess can react with the product). The mixture was stirred for 30 min
P{ H} NMR (121 MHz, CD
2
Cl
2
, 258C, H
3
PO
4
): d ˆ 13.3 (d, J(P,P) ˆ
2
13
at � 788C, and the cold bath was allowed to warm. After 4 ± 6 h, the
13 Hz, PPh
3
), � 5.5 (d, J(P,P) ˆ 13 Hz, PPh
2
H). The C NMR spectrum
‡
�
[13a]
volatiles were removed by rotary evaporation. The red foam was dried
was similar to that of (S)-2 OTs .
�
3
(
10 Torr, 6 h) to give (S)-5 (0.520 g, 0.550 mmol, 89%), which was pure by
NMR. A portion (0.035 g, 0.037 mmol) was dissolved in benzene, and
hexanes were added by vapor diffusion. Red prisms slowly formed, which
5
)] [(S)-3]:[19] This compound was pre-
2
(
S)-[(h -C
5
H
5
)Re(NO)(PPh
3
)(PPh
�
4
in 89 ± 96% yields by a deprotonation analogous to
‡
pared from (S)-2 BF
‡
� [13a]
that used to synthesize racemic 3 from 2 OTs .
M.p. 220 ± 2308C
6 6 0.5
were collected by filtration and dried as above to give (S)-5 ´ (C H )
decomp; [a] 52 28 9 ˆ 2058 (c ˆ 1.05 mgmL , THF).
� 1
[19]
1
13
31
(
0.020 g, 0.021 mmol, 60%). The IR and NMR ( H, C, P) spectra were
2
2
� 1
5
similar to those of the racemate. [a] ˆ 2168 (c ˆ 1.0 mg mL , THF).
[
(h -C
5
H
4
PPh
2
)Re(NO)(PPh
3
)(H)] (14)
589
5
‡
� �
‡
A) A Schlenk flask was charged with PPh
and THF (5 mL), and fitted with a septum. Then nBuLi (1.40m in hexane;
.580 mL, 0.790 mmol) was added with stirring. The colorless solution
2
H (0.13 mL, 0.14 g, 0.75 mmol)
[(h -C
5
H
4
PPh
2
)Re(NO)(PPh
3
)(m-PPh
2
)Rh(NBD)] PF
6
[(6 PF )]
6
6 6
A) A Schlenk flask was charged with 5 ´ (C H )
0.5 (0.505 g, 0.533 mmol) and
0
THF (25 mL), and [Rh(NBD)Cl] (0.12 g, 0.26 mmol) was added with
stirring. The orange solution turned deep red, and AgPF (0.155 g,
.582 mmol) was added. The sample became heterogeneous and red-
2
31
turned orange. A P NMR spectrum of an aliquot showed the clean
formation of LiPPh
(d ˆ � 29.90, s). The solution was transferred by
cannula with stirring to a septum-capped Schlenk tube that had been
6
2
0
brown. After 30 min, the volatiles were removed in vacuo. The residue was
extracted with benzene. The extract was filtered through a Celite plug. The
deep red filtrate was concentrated, and hexanes were added. The dark
orange solid was dissolved in a minimum of THF, and pentane was added
by vapor diffusion at � 208C. Orange-red, plate-like crystals slowly formed,
5
[30]
charged with [(h -C
5
H
5
)Re(NO)(PPh
3
)(OTs)] (13;
0.50 g, 0.70 mmol)
and THF (20 mL). The red solution was stirred and gradually turned
yellow. A 31P NMR spectrum of an aliquot showed complete product
formation. After 4 h, the volatiles were removed by oil-pump vacuum. The
yellow residue was extracted with benzene. The extract was filtered
through a Celite plug. The bright yellow filtrate was concentrated, and
hexanes added by vapor diffusion. Yellow needles slowly formed, which
were collected by filtration and dried by oil pump vacuum to give 14 (0.43 g,
�
3
which were collected by filtration and dried (10 Torr, 24 h) to give
‡
�
6
PF
6
´ THF (0.605 g, 0.445 mmol, 83%). M.p. 180 ± 1838C decomp;
NOP ReRh ´ C O (1325.1): C
elemental analysis (%) calcd for C54
H
47
F
6
4
4 8
H
�
1
5
1
(
7
2.57, H 4.48, P 9.35; found: C 52.53, H 4.52, P 9.34; IR (KBr, cm ): nÄ ˆ
0.60 mmol, 85%). M.p. 198 ± 2008C decomp; elemental analysis calcd (%)
‡
670 (s, NO); MS (FAB, 3-NBA): m/z (%): 1108 (58) [M] , 600 (50), 183
for C35 Re for: C 57.68, H 4.15, P 8.50; found: C 57.62, H 4.35, P
H
30NOP
2
1
74), 154 (100); H NMR (300 MHz, CD
.70 ± 7.65, 7.51 ± 7.43, 7.36 ± 7.31, 7.29 ± 7.23, 7.23 ± 7.08, 6.71 ± 6.64 (7m,
2
Cl
2
, 288C, TMS): d ˆ 8.10 ± 8.04,
�
1
1
8
CD
.34; IR (KBr, cm ): nÄ ˆ 1950 (s, ReH), 1633 (s, NO); H NMR (300 MHz,
2
Cl
2
, 288C, TMS): d ˆ 7.50 ± 7.31 (brm, 5C
6
H
5
), 4.91, 4.87, 4.65, 4.58
7
C
6
H
5
), 5.69, 5.51, 5.45, 4.66, 4.52, 4.45, 4.01, 4.00, 3.79, 3.69 (10brs,
, NBD CH), 3.69 ± 3.65, 1.84 ± 1.79 (2m, THF), 1.56 (dd, J(H,H') ˆ
2
3
(
4brm, C
5
H
4
), � 9.56 (dd, J(H,P) ˆ 29.7 Hz, J(H,P) ˆ 1.8 Hz, ReH);
2
C
5
H
4
1
3
1
C{ H} NMR (75 MHz, CD
2
Cl
2
, 288C, TMS): PPh
3
at d ˆ 138.2 (d,
at 138.7
3
2
8
.9 Hz, J(H,H'') ˆ 1.6 Hz, NBD-CHH'), 1.44 (dd, J(H',H) ˆ 8.5 Hz,
1
2
J(C,P) ˆ 53 Hz, i), 134.0 (d, J(C,P) ˆ 11 Hz, o), 129.0 (s, p); PPh
2
2
3
13
1
J(H',H'') ˆ 1.5 Hz, NBD-CHH'); C{ H} NMR (75 MHz, CD
2
Cl
2
, 288C,
1
1
(
d, J(C,P) ˆ 57 Hz, i), 138.5 (d, J(C,P) ˆ 56 Hz, i'), 133.7 (d, J(C,P) ˆ
1
1
TMS): PPh
3
, PPh
2
at d ˆ 142.0 (d, J(C,P) ˆ 20 Hz, i), 134.2 (d, J(C,P) ˆ
9
Hz, o), 130.4 (s, p); other PPh
3
, PPh
2
at 128.7 ± 128.5 (m); C
5
H
4
at 94.9 (d,
1
5
4 Hz, i'), 137.3 (d, J(C,P) ˆ 20 Hz, i''), 136.2 (d, J(C,P) ˆ 14 Hz), 134.8 (d,
1
2
J(C,P) ˆ 14 Hz, CP), 93.3 (s), 89.9 (d, J(C,P) ˆ 10 Hz), 88.9 (s), 88.7 (s);
J(C,P) ˆ 13 Hz), 133.6 (s), 133.6 (d, J(C,P) ˆ 11 Hz), 131.8 (d, J(C,P) ˆ
3
1
1
P{ H} NMR (121 MHz, CD
s, C PPh ).
2
Cl
2
, 258C, H
3
PO
4
): d ˆ 26.1 (s, PPh ), � 14.9
3
1
(
1 Hz), 131.2 ± 130.9 (m), 130.1 ± 128.8 (m); C
5
H
4
, NBD at 114.6 (s), 114.5
(
5
H
4
2
1
s), 110.8 (s), 110.2 (s), 102.5 (d, J(C,P) ˆ 18 Hz, CP), 98 (s), 96.1 (brm),
B)[19] An analogous synthesis was conducted with (S)-13 (0.200 g,
89.1 (brm), 83.2 (s), 81.7 (brm), 71.8 (br m), 69.9 (s); 68.2, 26.0 (2s, THF);
.280 mmol).[ A similar workup gave yellow needles of (R)-14 (0.160 g,
30]
31
P{ H} NMR (121 MHz, CD
1
Cl
, 288C, H
PO
): d ˆ 50.4 (ddd, J(P,Rh) ˆ
1
0
0
2
2
3
4
1
13
31
2
3
2
.220 mmol, 79%). The IR and NMR ( H, C, P) spectra were similar to
183 Hz, J(P,P)ˆ 19 Hz, J(P,P)ˆ 5 Hz, C
5
H
4
PPh
2
), 9.8 (dd, J(P,P) ˆ 14 Hz,
2024
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2024 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

ªChiral-at-Rheniumº Complexes
2015±2027
3
2
1
2
3
J(P,P) ˆ 5 Hz, PPh
3
), � 49.2 (ddd, J(P,Rh) ˆ 127 Hz, J(P,P) ˆ 19 Hz,
NMR (161.7 MHz, CDCl
3
, 288C, H
3
PO
4
): d ˆ 8.1 (d, J(P,P) ˆ 8 Hz) or 8.1
1
3
2
3
J(P,P) ˆ 14 Hz, RePPh
2
), � 144.0 (sep, J(P,F) ˆ 708 Hz, PF
6
).
(dd, J(P,P) ˆ 6 Hz, J(H,P) ˆ 12.1 Hz) (PPh
PPh ) or 25.7 (brs, PPh ).
2
), 25.8 (d, J(P,P) ˆ 8 Hz,
B) A sample was dissolved in CHCl
3
and layered with hexanes. Dark red
formed, and were used for crystallography
3
3
‡
�
‡
�
prisms of 6 PF
below). The solvate was verified by C NMR.
C)[19] The compounds (S)-5 ´ (C
0.5 (0.185 g. 0.192 mmol), THF (15 mL),
Rh(NBD)Cl] (0.046 g, 0.096 mmol), and AgPF (0.50 g, 0.200 mmol) were
6
´ (CHCl
3
)
2
B) An analogous synthesis was conducted with (S)-8 BF4 (1.420 g,
1.710 mmol). Workup gave (S)-9 ´ C H6 as orange needles (1.246 g,
1
3
(
6
1.518 mmol, 89%). M.p. 1728C decomp; [a]24
ˆ 2208 (c ˆ 2.70 mgmL
� 1
,
H
6
)
589
6
2 6 6
THF); elemental analysis (%) calcd for C36H32NOP Re ´ C H (820.9): C
[
2
6
1
6
1.45, H 4.67, N 1.71; found: C 61.15, H 4.68, N 1.71. The NMR spectra ( H,
C, P) were similar to those of the racemate. The crystallization
supernatant was kept at room temperature for several hours. Clear orange
cubes (0.2 ± 1.0 mm edges) of (S)-9 ´ C H6 formed. One was removed for a
crystal structure (below). The supernatant was decanted. The remaining
cubes were dried under a N stream (0.065 g, 0.079 mmol, 5%). Elemental
analysis (%) found: C 61.86, H 4.66, N 1.75 (calcd, see above).
combined as in procedure A. After 2 h, the volatiles were removed in
vacuo. The residue was extracted with benzene. The extract was filtered
through a Celite plug. The solvent was removed from the filtrate by rotary
evaporation. The residue was dried (10 Torr, 6 h) to give (R)-6 PF
THF (0.24 g, 0.18 mmol, 92%) as a dark orange powder that was pure by
1
3
31
�
3
‡
�
´
6
6
[19]
1
13
31
NMR ( H, C, P; data similar to racemate). Crystallization attempts gave
2
oils. [a] 52 28 9 ˆ 488 (c ˆ 1.0 mg mL� 1, THF).
5
5
‡
�
‡
�
[(h -C H PPh )Re(NO)(PPh )(CH PPh )] (11)
[
(h -C
5
H
5
)Re(NO)(PPh
3
)(CH
2
PPh
2
H)] BF
4
[8 BF
4
]
5
4
2
3
2
2
A) A Schlenk flask was charged with 1 (1.000 g, 1.790 mmol) and CH
2
Cl
2
�
A) A Schlenk tube was charged with 9 (1.210 g, 1.629 mmol) and THF
‡
(60 mL), and was cooled to � 608C (acetone/N slurry). A solution of tBuLi
(
(
50 mL), and was cooled to � 608C (acetone/N
2
slurry). Then Ph
3
C BF
4
2
0.650 g, 1.97 mmol, 1.1 equiv) was added with stirring. Within 30 min, the
(1.5m in pentane; 1.30 mL, 1.96 mmol, 1.2 equiv) was slowly added against
orange suspension turned to a light green-yellow solution. Then PPh
2
H
a N
orange mixture turned orange-red. An aliquot was assayed by P NMR
(Table 1; complete formation of 10). After 30 min, PPh Cl (0.331 mL,
2
flow with stirring. The cold bath was replaced by a 08C ice bath. The
31
(
0.0168 mL, 0.180 g, 0.967 mmol, 1.2 equiv) was added dropwise with
stirring. After 10 min, the cold bath was removed. The solution turned
orange and then red. After 1.5 h, the mixture was concentrated to about
2
0.395 g, 1.792 mmol) was added. The bath was allowed to warm to room
temperature over the course of 1 h. The solvent was removed by oil-pump
vacuum. Benzene (20 mL) was added. The mixture was filtered through a
Celite plug (2 Â 6 cm; with benzene rinses). The filtrate was concentrated
to 10 mL. A pentane layer (30 mL) was gently added. After 2 d, the
supernatant was decanted from a mixture of bright red crystals and yellow
powder to give 11 (1.024 g, 1.105 mmol, 68%). M.p. 115 ± 1188C decomp;
elemental analysis (%) calcd for C H NOP Re (927.0): C 62.19, H 4.46, N
15 mL by a brief exposure to oil pump vacuum. Some product crystallized.
The sample was layered with hexanes (40 mL). After 24 h, the orange-red
prisms were collected by filtration, washed with hexanes (2 Â 5 mL), and
�
3
‡
�
4
2 2
dried (10 Torr, 1 h) to give 8 BF ´ CH Cl (1.560 g, 1.704 mmol, 95%).
4 2
M.p. 202 ± 2068C; elemental analysis (%) calcd for C36H33BF NOP Re ´
CH Cl (915.6): C 48.54, H 3.85, N 1.53; found: C 48.65, H 3.87, N 1.54; IR
2
2
�
1
‡
(
5
CD
4
KBr, cm ): nÄ ˆ 1662 (s, NO); MS (FAB, 3-NBA): m/z (%): 744 (87) [M] ,
48 41
3
‡
‡
1
1.51; found: C 62.02, H 4.81, N 1.14; IR (KBr, cm� 1): nÄ ˆ 3051, 2907, 2868 (w,
58 (100) [M � HPPh
2
] , 481 (24) [M � PPh
3
] ; H NMR (400 MHz,
), 7.18 (ddd, 1J(H,P) ˆ
CH), 1637 (s, NO); MS (FAB, 3-NBA): m/z (%): 926 (38) [M]
‡
, 742 (90)
, 665 (50)
2
Cl
2
, 288C, TMS): d ˆ 7.86 ± 7.25 (m, 5C
6 5
H
3
3
[M � PPh ]
‡
‡
, 727 (35) [M � CH PPh ]
‡
, 681 (100) [M � OPPh ]
‡
89 Hz, J(H,H) ˆ 11.0 Hz, J(H,H') ˆ 5.1 Hz, HP), 5.32 (s, CH
2
Cl
Cl
2
), 4.95
2
2
2
3
1
3
1
[M � PPh ]
;
1
H NMR (400 MHz, [D ]THF, 288C, TMS): d ˆ 7.56 ± 7.02 (m,
(
s, C
5
H
5
), 2.58 ± 2.44 (m, CHH'); C{ H} NMR (100.5 MHz, CD
2
2
, 288C,
3
8
1
2
7C H ), 5.22, 4.82, 4.70, 3.39 (4brs, C H ), 2.41 (dd,
2
J(H,H') ˆ 11.8 Hz,
TMS): PPh
3
at d ˆ 134.3 (d, J(C,P) ˆ 55 Hz, i), 133.9 (d, J(C,P) ˆ 11 Hz,
6
5
5
4
3
J(H,P) ˆ 9.6 Hz, CHH'), 1.88 (dd,
2
J(H',H) ˆ 11.8 Hz, J(H',P) ˆ 1.9 Hz,
H}NMR (100.6 MHz, [D ]THF, 288C, TMS): PPh at d ˆ 136.6
o), 131.4 (s, p), 129.3 (d, J(C,P) ˆ 9 Hz, m); PPhPh' at 134.5 (s, p), 132.5 (d,
2
2
3
CHH'); 13C{
1
J(C,P) ˆ 9 Hz, o), 132.0 (d, J(C,P) ˆ 9 Hz, o'), 130.3 (d, J(C,P) ˆ 11 Hz,
8
3
3
1
(d,
1
J(C,P) ˆ 52 Hz, i), 134.7 (d,
J(C,P) ˆ 13 Hz, m); 2PPhPh' at 148.0 (d,
J(C,P) ˆ 20 Hz, i'), 139.4 (d, J(C,P) ˆ 13 Hz, i''), 137.7 (d,
2
J(C,P) ˆ 11 Hz, o), 130.6 (s, p), 129.0 (d,
J(C,P) ˆ 22 Hz, i), 146.7 (d,
J(C,P) ˆ 11 Hz,
m), 130.1 (d, J(C,P) ˆ 11 Hz, m'), 124.7 (d, J(C,P) ˆ 72 Hz, i), 123.3 (d,
1
1
31
1
3
1
J(C,P) ˆ 85 Hz, i'); 91.0 (s, C
5
H
5
), � 35.4 (d, J(C,P) ˆ 29 Hz, CH
2
); P{ H}
3
1
3
1
1
1
and P NMR (161.7 MHz, CD
2
Cl
2
, 288C, H
), 30.2 (d, J(P,P) ˆ 12 Hz) or 30.2 (d,
3
PO
4
): d ˆ 21.7 (d, J(P,P) ˆ
3
i'''), 134.8 (s, p), 128.0 (s, p'), 134.2 (d,
2
J(C,P) ˆ 20 Hz, o), 133.6 (d,
2
1
2 Hz) or 21.7 (brs, w1/2 ˆ 38 Hz) (PPh
3
1
2
J(C,P) ˆ 18 Hz, o'), 133.5 (d,
2
J(C,P) ˆ 15 Hz, o''), 129.5 (d,
J(C,P) ˆ
J(H,P) ˆ 487 Hz, each line with n1/2 ˆ 42 Hz) (PPh
2
).
3
3
1
1
9
5 Hz, o'''), 129.1 (d, J(C,P) ˆ 4 Hz, m), 128.1 (d, J(C,P) ˆ 4 Hz, m'),
B) An analogous synthesis was conducted with (S)-1 (1.000 g,
.790 mmol).[9] The CH
Cl solution (15 mL) was layered with pentane
40 mL). After 1 d, red prisms were collected by filtration, washed with
3
27.4 (d, J(C,P) ˆ 7 Hz, m''); C
5
H
4
at 105.5 (brs), 98 (d, J(C,P) ˆ 17 Hz),
1
2
2
1.8 (d, J(C,P) ˆ 4 Hz), 91.2 (s), 89.1 (d, J(C,P) ˆ 18 Hz, CP); � 18.2 (d,
(
1
31 1
J(C,P) ˆ 37 Hz, CH
2
); P{ H}NMR (161.7 MHz, [D
8
]THF/C
6
D
6
, 288C,
‡
�
pentane (2 Â 5 mL), and dried by oil pump vacuum to give (S)-8 BF
4
3
3
H
3
PO
4
): d ˆ 26.3/26.3 (d, J(P,P) ˆ 5/8 Hz, PPh
3
), 7.3/6.9 (dd, J(P,P) ˆ 5/3,
(
1.460 g, 1.758 mmol, 98%). M.p. 192 ± 1968C; elemental analysis (%) calcd
3
5
/8 Hz, C
5
H
4
PPh
2
), � 17.3/ � 17.7 (d, J(P,P) ˆ 5/3 Hz, PPh
2
).
(0.780 g, 1.050 mmol) and
THF (30 mL), and cooled to � 308C (acetone/N slurry). Then tBuLi
(1.50m in pentane, 1.05 mL, 1.58 mmol, 1.5 equiv) and PPh Cl (0.272 mL,
.324 g, 1.47 mmol, 1.4 equiv) were added as in procedure A. After 1 h, the
solvent was removed by oil pump vacuum. Benzene was added (10 mL).
The mixture was filtered through Celite plug. The filtrate was
concentrated to 5 mL. A pentane layer (10 mL) was gently added. After
d, the supernatant was decanted. The residue was washed with pentane
for C36
H
33BF
4
NOP
2
Re (830.6): C 52.06, H, 4.00, N 1.69; found: C 52.02, H
2
4
� 1
4
.08, N 1.55; [a]589 ˆ 1758 (c ˆ 1.68 mgmL , CHCl
3
). The NMR spectra
6 6
B) A Schlenk flask was charged with (S)-9 ´ C H
1
13
31
(
H, C, P) were similar to those of the racemate.
2
5
2
[
(h -C
5
H
5
)Re(NO)(PPh
3
)(CH
2
PPh
2
)] (9)
0
‡
�
A) A Schlenk tube was charged with
8
BF
4
´ CH
2
Cl
2
(1.554 g,
1.697 mmol) and THF (60 mL). A solution of tBuOK (1.0m in THF;
a
2.43 mL, 2.43 mmol) was added with stirring. After 1 h, the solvent was
removed by oil pump vacuum. Benzene (20 mL) was added, and the sample
2
was filtered through a Celite plug (4 Â 2 cm). The filtrate was concentrated
and dried by oil pump vacuum. The supernatant was evaporated to dryness
and the precipitation repeated. The two crops were combined to give (S)-11
as a orange powder (0.677 g, 0.730 mmol, 70%). Elemental analysis (%)
(
to ca. 10 mL) and layered with pentane (30 mL). After 24 h, the
supernatant was decanted from orange-red needles, which were dried by
oil pump vacuum to give 9 (1.250 g, 1.548 mmol, 90%). M.p. 178 ± 1798C
calcd for C48
H
41NOP
3
Re (927.0): C 62.19, H 4.46, N 1.51; found: C 61.73, H
decomp; elemental analysis (%) calcd for C36
H
32NOP
2
Re (742.8): C 58.21,
24
� 1
1
4
.60, N 1.41; [a]589 ˆ 1308 (c ˆ 2.80 mg mL , THF). The NMR spectra ( H,
�
1
H 4.34, N 1.89; found: C 58.32, H 4.25, N 1.68; IR (KBr, cm ): nÄ ˆ 3051 (m,
13
31
C, P) were similar to those of the racemate.
‡
CH), 1638 (s, NO); MS (FAB, 3-NBA): m/z (%): 742 (40) [M] , 558 (100)
5
‡
�
‡
�
‡
‡
1
[(h -C
5
H
4
PPh
2
)Re(NO)(PPh
3
)(m-CH
2
PPh
2
)Rh(NBD)] PF
6
(12 PF
6
)
[
MH � PPh
2
] , 481 (66) [M � PPh
3
] ; H NMR (400 MHz, CDCl , 288C,
3
TMS): d ˆ 7.62 ± 7.16 (m, 5C
6
H
5
), 4.86 (s, C
5
H
5
), 2.49 (dd, J(H,P) ˆ 9.9 Hz,
A) A Schlenk tube was charged with 11 (1.024 g, 1.105 mmol) and THF
(100 mL), and [Rh(NBD)Cl] (0.255 g, 0.552 mmol) was added with
stirring. After 30 min, AgPF (0.279 g, 1.105 mmol) was added. The sample
2
2
J(H,H') ˆ 12.1 Hz, CHH'), 1.84 (dd, J(H',P) ˆ 2.0 Hz, J(H',H) ˆ 12.1 Hz,
2
13
1
CHH'); C{ H} NMR (100.4 MHz, CDCl
3
, 288C, TMS): PPh
3
at d ˆ 135.8
6
1
2
(
d, J(C,P) ˆ 53 Hz, i), 133.6 (d, J(C,P) ˆ 11 Hz, o), 130.1 (s, p), 128.4 (d,
became heterogeneous and deep brown. After 2 h, the volatiles were
removed by oil pump vacuum. Benzene (60 mL) was added, and the
mixture was filtered through a Celite plug (3 Â 5 cm). The solvent was
removed from the filtrate by rotary evaporation. The residue was dried by
3
1
J(C,P) ˆ 9 Hz, m); PPhPh' at 146.6 (d, J(C,P) ˆ 20 Hz, i), 145.3 (d,
1
2
2
J(C,P) ˆ 18 Hz, i'), 133.0 (d, J(C,P) ˆ 18 Hz, o), 132.7 (d, J(C,P) ˆ 17 Hz,
3
3
o'), 127.7 (d, J(C,P) ˆ 7 Hz, m), 127.6 (d, J(C,P) ˆ 6 Hz, m'), 127.4 (s, p),
1
31
1
31
‡
�
1
27.0 (s, p'); 89.8 (s, C
5
H
5
), � 19.5 (d, J(C,P) ˆ 35 Hz, CH
2
); P{ H} and
P
oil pump vacuum to give crude 12 PF
6
(1.330 g, 1.050 mmol, 95%) as a
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2025 $ 17.50+.50/0
2025

J. A. Gladysz et al.
FULL PAPER
reddish brown solid. A sample (0.110 g, 0.087 mmol) was dissolved in THF
In another series of determinations, 17a,b (ca. 0.010 g in 2 mL methanol)
were treated with diazomethane/diethyl ether (yellow endpoint) to give
methyl esters (solvent was removed under vacuum, the residue was
extracted with HPLC grade isopropanol, and the extract filtered through
(
10 mL). The solution was concentrated to ca. 5 mL, and pentane (15 mL)
was added. The solid was collected on a frit, washed with small amounts of
‡
�
6
as a light brown
pentane, and dried by oil pump vacuum to give 12 PF
powder that was pure by NMR (0.065 g, 0.051 mmol, 59%). M.p. 180 ±
858C decomp; elemental analysis (%) calcd for C55 NOP ReRh
1267.0): C 52.14, H 3.90, N 1.11; found: C 51.72, H 4.31, N 0.83; IR (KBr,
glass wool). The ester from 17a was analyzed by GLC (1008C, N
2
carrier
�
1
1
(
H
49
F
6
4
flow 20 mLmin , 25 m  0.4 mm glass capillary column packed with a
[47]
modified b-cyclodextrin on silica) to give the data in entries 1 and 5. The
ester from 17b, and esters 17c,d, were analyzed by HPLC (typically 98:2 v/v
isohexane/isopropanol (isocratic), Chiralcel OD with cellulose-carbamate
on silica gel) to give the data in entries 6 ± 15.
�
1
cm ): nÄ ˆ 3056, 2924 (w, CH), 1663 (s, NO), 1481 (m), 1435 (m), 1309 (w),
1
261 (w), 1186 (w), 1160 (w), 1094 (m), 1027 (w), 999 (w), 839 (s, PF), 744
‡
(
m), 696 (m); MS (FAB, 3-NBA): m/z (%): 1122 (100) [M] , 1030 (40)
‡
1
[
M � NBD] ; H NMR (400 MHz, CDCl
3
, 288C, TMS): d ˆ 7.59 ± 7.01 (m,
), 4.90, 4.53, 4.43, 4.00, 3.95, 3.72
), 2.25 (m, CHH'), 2.07 (m, CHH');
Crystallography: Data were collected as summarized in Table 2. Cell
7
C
6
H
5
), 5.60, 5.42, 4.86, 4.00 (4brs, C
5
H
4
‡
�
6 3 2
´ (CHCl ) were determined from 15 reflections
parameters for 6 PF
168 < 2q < 298). Lorentz-polarization corrections were applied. The struc-
(
6brs, NBD-CH), 1.42 (brs, NBD-CH
2
(
1
3
1
C{ H} NMR (100.6 MHz, CDCl
3
, 288C, TMS): d ˆ PPh
3
at 133.1 (d,
ture was solved by standard heavy atom techniques (all data by full-matrix-
1
2
4
J(C,P) ˆ 53 Hz, i), 132.5 (d, J(C,P) ˆ 10 Hz, o), 129.8 (d, J(C,P) ˆ 2 Hz,
[48a]
least-squares on F) using the UCLA crystallographic package.
Carbon
3
1
p), 127.8 (d, J(C,P) ˆ 11 Hz, m); 2 RhPPhPh' at 138.2 (d, J(C,P) ˆ 29 Hz,
atoms were refined isotropically, and hydrogen atom positions were
1
2
i), 134.1 (d, J(C,P) ˆ 47 Hz, i'), 134.1 (d, J(C,P) ˆ 11 Hz, o), 133.4 (d,
calculated. Other atoms were refined anisotropically (D/d (max) ˆ 2.40).
2
2
2
J(C,P) ˆ 13 Hz, o'), 131.7 (d, J(C,P) ˆ 11 Hz, o''), 131.2 (d, J(C,P) ˆ
‡
�
6
6 6 2 2
Cell parameters for (S)-9 ´ C H and 12 PF ´ CH Cl were determined
3
3
1
0 Hz, o'''), 130.1 (d, J(C,P) ˆ 8 Hz, m), 128.4 (d, J(C,P) ˆ 10 Hz, m'),
from 15 reflections (5.08 < 2q < 50.08). Lorentz-polarization and empirical
absorption (Y scans) corrections were applied. Space groups were
determined from systematic absences and subsequent least-squares refine-
ment. The structures were solved by direct methods. The data were refined
1
28.0 (d, 3J(C,P) ˆ 11 Hz, m''), 127.6 (d, J(C,P) ˆ 7 Hz, m''') 130.8 (d,
3
4
4
4
J(C,P) ˆ 2 Hz, p), 129.9 (d, J(C,P) ˆ 3 Hz, p'), 129.7 (d, J(C,P) ˆ 2 Hz,
4
p''), 128.3 (d, J(C,P) ˆ 2 Hz, p'''); C
5 4
H and NBD at 97.6 (s), 94.8 (d,
1
J(C,P) ˆ 16 Hz, CP), 93.8 (s), 90.5 (m), 87.9 (m), 85.2 (s), 84.3 (s), 80.2 (s),
2
[48b]
(
all data by full-matrix-least-squares on F ) using SHELXL-93.
Non-
3
1
1
6
9.1 (s), 67.9 (s), 53.7 (s), 52.5 (s); � 14.2 (brs, ReCH
2
); P{ H} NMR
hydrogen atoms were refined with anisotropic thermal parameters. Hydro-
gen atoms were fixed in idealized positions using a riding model. Scattering
1
(
161.7 MHz, CDCl
3
, 288C, H
3
PO
4
): d ˆ 50.5 (ddd, J(P,Rh) ˆ 148 Hz,
2
3
1
J(P,P) ˆ 34 Hz, J(P,P) ˆ 18 Hz, C
5
H
4
PPh
PPh
2
), 23.9 (ddd, J(Rh,P) ˆ 166 Hz,
[49]
factors were taken from literature. The rhenium configuration in (S)-9 ´
2
3
3
3
J(P,P) ˆ 34 Hz, J(P,P) ˆ 4 Hz, CH
2
2
), 20.2 (dd, J(P,P) ˆ 18 Hz,
C
6
H
6
was established by Flackꢁs x parameter (found: � 0.006(12); theory
1
J(P,P) ˆ 4 Hz, PPh
3
), � 156.5 (sep, J(P,F) ˆ 708 Hz, PF
6
).
[50]
for correct and inverted structures: 0 and 1).
‡
�
B) A solution of crude 12 PF
with hexane (30 mL). After three weeks, deep red prisms of 12 PF
CH Cl formed. One was removed for a crystal structure (below). The
supernatant was decanted, and the remaining prisms were dried under a N
stream. Elemental analysis (%) calcd for C56 NOP ReRh (1351.9):
C 49.75, H 3.80, N 1.04; found: C 49.72, H 3.97, N 0.97.
C) A Schlenk tube was charged with (S)-11 (0.609 g, 0.657 mmol) and THF
50 mL), and [Rh(NBD)Cl] (0.151 g, 0.328 mmol) was added with stirring.
After 1 h, AgPF (0.166 g, 0.657 mmol) was added. After 1 h, the volatiles
6
(0.070 g) in CH
2
Cl
2
(5 mL) was layered
Crystallographic data (excluding structure factors) for the structures
reported in this paper have been deposited with the Cambridge Crystallo-
graphic Data Centre as supplementary publication no. refcode FOWNIO
‡
�
6
´
2
2
‡
�
‡
�
2
[
6
PF
´ (CHCl
3 2 6 6
) ], CCDC-147776 [(S)-9 ´ C H ] and -147777 [12 PF ´
6
6
H
51Cl
2
F
6
4
CH
2
Cl
2
]. Copies of the data can be obtained free of charge on application
to CCDC, 12 Union Road, Cambridge CB21EZ, UK (fax: (‡44)1223-336-
033; e-mail: deposit@ccdc.cam.ac.uk).
(
2
6
were removed by oil pump vacuum. Benzene (30 mL) was added. The
mixture was filtered through a Celite plug. The solvent was removed from
the filtrate by oil pump vacuum. The brown semisolid was dissolved in a
minimum of benzene, and pentane was added. The precipitate was
collected by filtration, washed with pentane (10 mL) and dried by oil
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft (DFG; GL 300-1/4 and
postdoctoral fellowship, O.M.) and US National Science Foundation for
support, and Johnson Matthey PMC for a loan of rhodium.
‡
�
pump vacuum to give (S)-12 PF
6
as a deep brown powder (0.680 g,
0
.537 mmol, 82%). This was reprecipitated from benzene/pentane to give a
2
4
red-brown powder. M.p. 180 ± 1858C decomp; [a]589 ˆ � 658 (c ˆ
[
1] a) Comprehensive Asymmetric Catalysis I ± III (Eds.: E. N. Jacobsen,
A. Pfaltz, H. Yamamoto), Springer, Berlin, Germany, 1999; b) Cata-
lytic Asymmetric Synthesis (Ed.: I. Ojima), 2nd ed., Wiley-VCH, New
York, 2000.
�
1
0
.80 mgmL , THF). Both samples showed small amounts of impurities
13
by NMR (<2%), and microanalyses were slightly off. The C NMR
spectra were similar to that of the racemate, but the H and 31P NMR
1
spectra showed minor differences. Hence, these data are given below.
[
2] K. Mu nÄ iz, C. Bolm, Chem. Eur. J. 2000, 6, 2309.
1
H NMR (see racemate): d ˆ 7.62 ± 7.05 (m, 7 C
4brs, C ), 4.84, 4.58, 4.50, 4.05, 3.96, 3.81 (6brs, NBD-CH), 1.47 (brs,
NBD-CH
d ˆ 50.7 (ddd, J ˆ 148, 34, 18 Hz, C
CH PPh
), 19.7 (dd, J ˆ 18, 4 Hz, PPh
6 5
H ), 5.46, 5.42, 4.80, 4.43
[3] Reviews, perspectives, and commercial applications: a) A. Togni,
Angew. Chem. 1996, 108, 1581; Angew. Chem. Int. Ed. Engl. 1996, 35,
(
5 4
H
3
1
1
2
), 2.30 (m, CHH'), 2.17 (m, CHH'); P{ H} NMR (see racemate):
PPh
), 22.6 (ddd, J ˆ 166, 34, 4 Hz,
), � 158.0 (sep, J ˆ 708 Hz, PF ).
1
2
475; b) C. J. Richards, A. J. Locke, Tetrahedron: Asymmetry 1998, 9,
377; c) A. Togni, N. Bieler, U. Burckhardt, C. Köllner, G. Pioda, R.
5
H
4
2
2
2
3
6
Schneider, A. Schnyder, Pure Appl. Chem. 1999, 71, 1531; d) H.-U.
Blaser, F. Spindler, Chapter 41.1 in ref. [1]; e) D. A. Dobbs, K. P. M.
Vanhessche, E. Brazi, V. Rautenstrauch, J.-Y. Lenoir, J.-P. Gen eà t, J.
Wiles, S. H. Bergens, Angew. Chem. 2000, 112, 2080; Angew. Chem.
Int. Ed. 2000, 39, 1992.
Hydrogenations (Table 5): A 50 mL flask was charged with 16 (3.00 mmol;
typical was entry 6, 0.388 g 16b), catalyst (0.50 mol%; entry 6: 0.019 g (S)-
‡
�
1
2
PF
orange solution was freeze-pump-thaw degassed (4 Â ). A H
was introduced, and the solution vigorously stirred. Within 1 min, H
uptake began. After H uptake ceased (entry 6: 58 mL, 2.6 mmol, theory:
6
), and THF (ca. 25 mL), and attached to a gas burette. The light
2
atmosphere
2
[
4] Lead references to ferrocene-based diphosphines with only planar
chirality elements: a) J. Kang, J. H. Lee, S. H. Ahn, J. S. Choi,
Tetrahedron Lett. 1998, 39, 5523; b) R. Kuwano, T. Uemura, M.
Saitoh, Y. Ito, Tetrahedron Lett. 1999, 40, 1327; c) M. T. Reetz, E. W.
Beuttenmüller, R. Goddard, M. Past o , Tetrahedron Lett. 1999, 40,
4977; d) G. Argouarch, O. Samuel, H. B. Kagan, Eur. J. Org. Chem.
2000, 2885; e) G. Argouarch, O. Samuel, O. Riant, J.-C. Daran, H. B.
Kagan, Eur. J. Org. Chem. 2000, 2893.
[5] a) H. Brunner, A. Winter, J. Breu, J. Organomet. Chem. 1998, 553,
285; b) T. Ohkuma, M. Kitamura, R. Noyori, Chapter 1 in ref. [1b].
[6] J. B. Lambert, Y. Zhao, R. W. Emblidge, L. A. Salvador, X. Liu, J.-H.
So, E. C. Chelius, Acc. Chem. Res. 1999, 32, 183; see the section
ªeffects beyond the b positionº.
2
6
2
7 mL), 17 was isolated by a standard workup (entry 6: 0.338 g per
.58 mmol 17b).[
27a]
Product configurations were assigned from the signs of optical rota-
tions.[
27a, 45]
Enantiomeric purities were assayed chromatographically. In
[46]
one series of determinations, 17a,b,d were first treated with methanol/
HCl. The resulting methyl esters, and 17c, were treated with trifluoroacetic
anhydride to give N-trifluoroacetyl-N-acetyl amino esters, which were
analyzed by GLC (1308C, N
column packed with 5% lauroyl-l-valine-tert-butylamide (Supelco SP300)
�
1
2
carrier flow 20 mLmin , 2 m  2 mm glass
[
18]
on 100/120 Supelcoport) to give the data communicated earlier and in
entries 2 ± 4, Table 5.
2026
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2026 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 9

ªChiral-at-Rheniumº Complexes
2015±2027
[
[
7] H. Brunner, Angew. Chem. 1999, 111, 1248; Angew. Chem. Int. Ed.
999, 38, 1194.
8] a) H. Brunner, J. Wachter, J. Schmidbauer, G. M. Sheldrick, P. G.
Strouse, J. A. Gladysz, Organometallics 1982, 1, 1204; c) G. L. Crocco,
K. E. Lee, J. A. Gladysz, Organometallics 1990, 9, 2819.
[30] J. H. Merrifield, J. M. Fern a ndez, W. E. Buhro, J. A. Gladysz, Inorg.
Chem. 1984, 23, 4022.
1
Jones, Angew. Chem. 1986, 98, 339; Angew. Chem. Int. Ed. Engl. 1986,
2
5, 371; H. Brunner, J. Wachter, J. Schmidbauer, G. M. Sheldrick, P. G
[31] G. L. Crocco, J. A. Gladysz, J. Am. Chem. Soc. 1988, 110, 6110.
[32] The following citations are restricted to reactions of neutral cyclo-
pentadienyl complexes and anionic species that give a neutral metal
hydride complex bearing a monosubstituted cyclopentadienyl ligand:
a) P. Brun, P. Vierling, J. G. Riess, G. Le Borgne, Organometallics
1987, 6, 1032; b) T. C. Forschner, J. A. Corella II, N. J. Cooper,
Organometallics 1990, 9, 2478, and earlier work from this group
therein.
Jones, Organometallics 1986, 5, 2212; b) H. Brunner, M. Prommes-
berger, Tetrahedron: Asymmetry 1998, 9, 3231.
[
9] F. Agbossou, E. J. OꢁConnor, C. M. Garner, N. Quir o s M e ndez, J. M.
Fern a ndez, A. T. Patton, J. A. Ramsden, J. A. Gladysz, Inorg. Synth.
1
992, 29, 211.
10] M. A. Dewey, Y. Zhou, Y. Liu, J. A. Gladysz, Organometallics 1993,
2, 3924.
[
[
1
11] a) D. M. Dalton, C. M. Garner, J. M. Fern a ndez, J. A.Gladysz, J. Org.
Chem. 1991, 56, 6823, and earlier full papers therein; b) Y. Wang, J. A.
Gladysz, J. Org. Chem. 1995, 60, 903; c) G. A. Stark, M. A. Dewey,
G. B. Richter-Addo, D. A. Knight, A. M. Arif, J. A. Gladysz in
Stereoselective Reactions of Metal-Activated Molecules (Eds.: H.
Werner, J. Sundermeyer), Vieweg, Braunschweig, Germany, 1995,
pp. 51; d) O. Meyer, P. C. Cagle, K. Weickhardt, D. Vichard, J. A.
Gladysz, Pure Appl. Chem. 1996, 68, 79.
12] J. A. Gladysz, B. J. Boone, Angew. Chem. 1997, 109, 566; Angew.
Chem. Int. Ed. Engl. 1997, 36, 550.
13] a) W. E. Buhro, B. D. Zwick, S. Georgiou, J. P. Hutchinson, J. A.
Gladysz, J. Am. Chem. Soc. 1988, 110, 2427; b) See also B. D. Zwick,
M. A. Dewey, D. A. Knight, W. E. Buhro, A. M. Arif, J. A. Gladysz,
Organometallics 1992, 11, 2673.
[33] One closely related example is the reaction of the iron bromide
5
complex [(h -C
5
H
5
)Fe(CO)(PPh(OEt)
2
)(Br)] and LiNEt
2
,
which
)-
5
gives the hydride complex [(h -C
5
H
4
NEt )Fe(CO)(PPh(OEt)
2
2
[
32a]
(H)].
A number of conceptually similar hydride migrations have
been reported. See for example R. P. Hughes, S. M. Maddock, A. L.
Rheingold, L. M. Liable-Sands, J. Am. Chem. Soc. 1997, 119, 5988.
[34] a) O. Meyer, A. M. Arif, J. A. Gladysz, Organometallics 1995, 14,
1844; b) T.-S. Peng, A. M. Arif, J. A. Gladysz, J. Chem. Soc. Dalton
Trans. 1995, 1857.
[35] a) S. Georgiou, J. A. Gladysz, Tetrahedron 1986, 42, 1109; b) S. G.
Davies, I. M. Dordor-Hedgecock, K. H. Sutton, M. Whittaker, J. Am.
Chem. Soc. 1987, 109, 5711.
[
[
[36] a) Conformations of ReCH
2
-XRR'R'' linkages have also been stud-
[
34]
ied.
Typically, the largest group (R'') gives a torsion angle of
[
14] a) Basicities of alkoxy ligands: S. K. Agbossou, J. M. Fern a ndez, J. A.
Gladysz, Inorg. Chem. 1990, 29, 476; b) nucleophilicities of amido
ligands: M. A. Dewey, D. A. Knight, A. M. Arif, J. A. Gladysz, Chem.
Ber. 1992, 125, 815.
15] F. B. McCormick, W. B. Gleason, X. Zhao, P. C. Heah, J. A. Gladysz,
Organometallics 1986, 5, 1778.
16] Other literature relevant to points discussed below: a) M. J. Burk, Acc.
Chem. Res. 2000, 33, 363; b) M. J. Burk, J. E. Feaster, W. A. Nugent,
R. L. Harlow, J. Am. Chem. Soc. 1993, 115, 10125; c) I. D. Gridnev, N.
Higashi, K. Asakura, T. Imamoto, J. Am. Chem. Soc. 2000, 122, 7183;
d) S. Feldgus, C. R. Landis, J. Am. Chem. Soc. 2000, 122, 12714;
e) C. R. Landis, J. Halpern J. Am. Chem. Soc. 1987, 109, 1746; f) W. Li,
X. Zhang, J. Org. Chem. 2000, 65, 5871, and earlier work from this
group therein.
approximately 1808, such that the X�R'' bond is anti to the Re�C
bond. Accordingly, (S)-9 exhibits a Re-C1-P2-C50 torsion angle of
� 176.98; b) for analyses of both conformational and configurational
equilibria in related amido and alkoxide complexes, see M. A. Dewey,
G. A. Stark, J. A. Gladysz, Organometallics 1996, 15, 4798. At
equilibrium, the smallest group (R) prefers the position analogous
to the phosphorus lone pair in crystalline (S)-9.
[
[
[37] a) T. Ireland, G. Groûheimann, C. Wieser-Jeunesse, P. Knochel,
Angew. Chem. 1999, 111, 3397; Angew. Chem. Int. Ed. 1999, 38,
3212; b) L. Schwink, P. Knochel, Chem. Eur. J. 1998, 4, 950.
[38] a) J. Kang, J. H. Lee, S. H. Ahn, J. S. Choi, Tetrahedron Lett. 1998, 39,
5523; b) J. J. A. Perea, A. Börner, P. Knochel, Tetrahedron Lett. 1998,
39, 8073.
5
[39] Many other phosphido complexes [(h -C
5
H
5
)Re(NO)(PPh
3
)(PRR')]
ligand
[
[
17] K. Kromm, J. A. Gladysz, unpublished results.
18] a) B. D. Zwick, A. M. Arif, A. T. Patton, J. A. Gladysz, Angew. Chem.
are described in ref. [13]. However, a replacement of the PPh
would necessitate a new resolution procedure.
3
1
987, 99, 921; Angew. Chem. Int. Ed. Engl. 1987, 26, 910; b) see also
[40] a) Y.-H. Huang, F. Niedercorn, A. M. Arif, J. A. Gladysz, J. Organo-
met. Chem. 1990, 383, 213; b) T.-S. Peng, C. H. Winter, J. A. Gladysz,
Inorg. Chem. 1994, 33, 2534.
[41] a) P. Sauvageot, O. Blacque, M. M. Kubicki, S. Jug e , C. Moïse,
Organometallics 1996, 15, 2399; b) C. Poulard, G. Boni, P. Richard, C.
Moïse, J. Chem. Soc. Dalton Trans. 1999, 2725.
[42] G. Helmchen, A. Pfaltz, Acc. Chem. Res. 2000, 33, 336.
[43] A. F. Burchat, J. M. Chong, N. Nielsen, J. Organomet. Chem. 1997, 542,
281.
[44] R. M. Herbst, D. Shemin in Organic Synthesis Coll., Vol. 2, Wiley,
New York, 1943, pp. 1.
[45] B. D. Vineyard, W. S. Knowles, M. J. Sabacky, G. L. Bachman, D. J.
Weinkauff, J. Am. Chem. Soc. 1977, 99, 5946.
[46] a) B. A. Andersson, Acta Chem. Scand. 1971, 25, 1514; b) Supelco
Technical Bulletin 765G; Supelco Inc, Bellefonte, PA; one of nine
references in this brief review: S. Nakaparskin, E. Gil-Av, J. Oro, J.
Anal. Biochem. 1970, 33, 374.
[47] A privately manufactured column was employed, but equal or better
separations would be realizable with commercial cyclodextrin fused
silica capillary columns (e.g. FS-LIPODEX; Macherey-Nagel).
[48] a) See footnote [49] in ref. [13a]; b) G. M. Sheldrick, SHELX-93,
University of Göttingen, 1993. See also G. M. Sheldrick in Crystallo-
graphic Computing 3 (Eds.: G. M. Sheldrick, C. Krüger, R. Goddard),
Oxford University, England, 1993, p. 175.
B. D. Zwick, Ph.D. Thesis, University of Utah, 1987.
19] The non-racemic five- and six-membered chelates were synthesized
from opposite enantiomers of 1. For ease of comparison, the
configurations of all complexes in the former series (and the hydro-
genation products) have been inverted in the text, Tables, Scheme and
Figures. The true configurations are designated in the Experimental
Section.
[
[
20] Chiral compounds not preceded by R/S descriptors are racemic.
Rhenium configurations are designated by a modified Cahn-Ingold-
[11]
Prelog system referenced in earlier papers. Priority sequence for
5
ligands in this study: (h -C
5
H
4
X) > PPh
> H.
21] J. A. Ramsden, C. M. Garner, J. A. Gladysz, Organometallics 1991, 10,
631.
22] J. J. Kowalczyk, S. K. Agbossou, J. A. Gladysz, J. Organomet. Chem.
990, 397, 333.
23] Abbreviations: OTs ˆ OSO
CF
2
Rh > PPh
3
> PPh
2
H > PPh
2
>
2 3
OTs > NO > CH P > CH
[
[
[
[
1
1
2
-p-C
6
H
5
CH
3
;
NBD ˆ norbornadiene;
OTf ˆ OSO
2
3
.
24] P. C. Cagle, O. Meyer, K. Weickhardt, A. M. Arif, J. A. Gladysz, J. Am.
Chem. Soc. 1995, 117, 11730.
[
[
25] J. J. Kowalczyk, A. M. Arif, J. A. Gladysz, Chem. Ber. 1991, 124, 729.
26] P. Johnston, M. S. Loonat, W. L. Ingham, L. Carlton, N. J. Coville,
Organometallics 1987, 6, 2121.
[
[
[
27] a) D. P. Riley, R. E. Shumate, J. Org. Chem. 1980, 45, 5187; b) R. R.
Schrock, J. A. Osborn, J. Am. Chem. Soc. 1971, 93, 2397.
28] J. H. Merrifield, G.-Y Lin, W. A. Kiel, J. A. Gladysz, J. Am. Chem.
Soc. 1983, 105, 5811.
29] a) W. Tam, G.-Y. Lin, W.-K. Wong, W. A. Kiel, V. K. Wong, J. A.
Gladysz, J. Am. Chem. Soc. 1982, 104, 141; b) J. Merrifield, C. E.
[49] D. T. Cromer, J. T. Waber in: International Tables for X-ray Crystal-
lography (Eds.: J. A. Ibers, W. C. Hamilton), Kynoch, Birmingham,
England, 1974.
[50] H. D. Flack, Acta Crystallogr. 1983, A39, 876.
Received: September 19, 2000 [F2742]
Chem. Eur. J. 2001, 7, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0709-2027 $ 17.50+.50/0
2027