Comparing chiral ferrocenyl and ruthenocenyl ligands: How subtle structural changes influence their performance in asymmetric catalysis

Article abstract of DOI:10.1021/om9508997

A route to enantiopure heteroleptic ruthenocenyl derivatives has been found; the diastereoselective addition of MeLi to (R)-CyCH(Me)N(Me)CH=C₅H₄ (de = 74%), followed by a transmetalation reaction with either [Cp*Ru(μ³-Cl)]₄ or [(p-cymene)RuCl₂]₂/KPF₆, afforded the heteroleptic complexes (S,R)-[Cp*Ru(η⁵-C₅H ₄CH(Me)N(Me)CH(Me)Cy)], 6, or (S,R)-[(p-cymene)Ru(η⁵-C₅H ₄CH(Me)N(Me)CH(Me)Cy)]⁺PF₆^-, 5, in 93% and 72% yields, respectively. Whereas 5 displayed a somewhat inert behavior, 6 reacted with NHMe₂, in acetic acid, to afford its dimethylamino congener (S)-7 in 93% yield. The latter was converted in two steps into the bis(phosphine) derivatives (S)-(R)-Cp*RuC₅H₃CH(Me)PCy₂PPh ₂-2, (S)-(R)-9, and (S)-(R)-Cp*RuC₅H₃CH(Me)PC₈H ₁₄PPh₂-2, (S)-(R)-10, and into the P,N derivative (S)-(R)-Cp*RuC₅H₃CH(Me){N₂C ₃HMe₂-3,5)PPh₂}-1,2, (S)-(R)-11. These products were obtained in >99% ee after recrystallization. The ruthenocenyl derivatives were probed for their use as chiral ligands for the palladium-catalyzed enantioselective allylic alkylation and the rhodium catalyzed hydroboration reactions. By employing the ruthenocenylpyrazole (S)-(R)-11, styrene was converted to (S)-1-phenylethanol with 87% ee, whereas its isostructural ferrocenyl congener (S)-(R)-16 afforded 94% ee. The following compounds were characterized by X-ray diffraction: (S,R)-5, (S)-(R)-9, (S)-(R)-10, [((S*)-(R*)-14)Pd(η³-C₃H₅H ₅)]⁺-[OTf]^- ((S*)-(R*)-12), and [((S*)-(R*)-9)Pd(η³-C₃H₅)] ⁺[OTf]^- ((S*)-(R*)-13).

Full text of DOI:10.1021/om9508997

1614
Organometallics 1996, 15, 1614-1621
Com p a r in g Ch ir a l F er r ocen yl a n d Ru th en ocen yl
Liga n d s: How Su btle Str u ctu r a l Ch a n ges In flu en ce
Th eir P er for m a n ce in Asym m etr ic Ca ta lysis
Hendrikus C. L. Abbenhuis,† Urs Burckhardt,† Volker Gramlich,‡
Arianna Martelletti,‡ J ohn Spencer,† Ivo Steiner,† and Antonio Togni*,†
Laboratory of Inorganic Chemistry and Institute of Crystallography and Petrography,
Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Zurich, Switzerland
Received November 21, 1995X
A route to enantiopure heteroleptic ruthenocenyl derivatives has been found; the
diastereoselective addition of MeLi to (R)-CyCH(Me)N(Me)CHdC5H4 (de ) 74%), followed
by a transmetalation reaction with either [Cp*Ru(µ3-Cl)]4 or [(p-cymene)RuCl2]2/KPF6,
afforded the heteroleptic complexes (S,R)-[Cp*Ru(η5-C5H4CH(Me)N(Me)CH(Me)Cy)], 6, or
(S,R)-[(p-cymene)Ru(η5-C5H4CH(Me)N(Me)CH(Me)Cy)]+PF6-, 5, in 93% and 72% yields,
respectively. Whereas 5 displayed a somewhat inert behavior, 6 reacted with NHMe2, in
acetic acid, to afford its dimethylamino congener (S)-7 in 93% yield. The latter was converted
in two steps into the bis(phosphine) derivatives (S)-(R)-Cp*RuC5H3CH(Me)PCy2PPh2-2, (S)-
(R)-9, and (S)-(R)-Cp*RuC5H3CH(Me)PC8H14PPh2-2, (S)-(R)-10, and into the P,N derivative
(S)-(R)-Cp*RuC5H3CH(Me){N2C3HMe2-3,5)PPh2}-1,2, (S)-(R)-11. These products were ob-
tained in >99% ee after recrystallization. The ruthenocenyl derivatives were probed for
their use as chiral ligands for the palladium-catalyzed enantioselective allylic alkylation
and the rhodium catalyzed hydroboration reactions. By employing the ruthenocenylpyrazole
(S)-(R)-11, styrene was converted to (S)-1-phenylethanol with 87% ee, whereas its isostruc-
tural ferrocenyl congener (S)-(R)-16 afforded 94% ee. The following compounds were
characterized by X-ray diffraction: (S,R)-5, (S)-(R)-9, (S)-(R)-10, [((S*)-(R*)-14)Pd(η3-C3H5)]+-
[OTf]- ((S*)-(R*)-12), and [((S*)-(R*)-9)Pd(η3-C3H5)]+[OTf]- ((S*)-(R*)-13).
Ch a r t 1
In tr od u ction
By virtue of their high scope of application as ligands
in a range of asymmetric reactions, optically active
ferrocenyl phosphines,1 exemplified by 1-3, are becoming
of increasing importance (see Chart 1).2 The facile
synthesis of a wide range of such derivatives, by a
nucleophilic substitution reaction occurring with reten-
tion of configuration at the stereogenic center of the side
chain, and the resulting structural diversity on the
ferrocenyl unit, allow for a systematic study, and
exploitation, of electronic and steric effects in enantio-
selective catalysis.2e
Recently, we reported a new stereoselective route to
congeners of 1-3, with the added advantage of allowing
variation of the (“lower”) nonfunctionalized Cp.3 The
key feature of the underlying synthetic strategy was to
construct ferrocenes starting from enantiomerically
enriched Cp synthons and a suitable iron(II) precursor.
Hence, the synthesis of precursors to analogues of 1-3
was possible via a highly stereoselective addition of
MeLi to one diastereotopic face of an enantiomerically
pure fulvene derivative, followed by transmetalation
with an iron(II) salt. The diastereoselectivity observed
during the formation of one particular metalated cyclo-
pentadienyl derivative (S,R)-4 was as high as 87%. As
a natural extension to this study, it was tempting to
test whether this method was amenable for the synthe-
sis of the corresponding heteroleptic ruthenocene com-
plexes, namely by reaction of 4 with appropriate ruthe-
nium(II) derivatives. This choice was potentially
† Laboratory of Inorganic Chemistry.
‡ Institute of Crystallography and Petrography.
X Abstract published in Advance ACS Abstracts, February 15, 1996.
(1) For reviews, see: (a) Hayashi, T. In Ferrocenes. Homogeneous
Catalysis, Organic Synthesis, Materials Science; Togni, A., Hayashi,
T., Eds.; VCH: Weinheim, Germany, 1995; pp 105-142. (b) Sawamura,
M.; Ito, Y. Chem. Rev. 1992, 92, 857-871. See also: (c) Hayashi, T.;
Mise, T.; Fukushima, M.; Kagotani, M.; Nagashima, N.; Hamada, Y.;
Matsumoto, A.; Kawakami, S.; Konishi, M.; Yamamoto, K.; Kumada,
M. Bull. Chem. Soc. J pn. 1980, 53, 1138-1151.
(2) For recent reports from our laboratories, see: (a) Togni, A.;
Breutel, C.; Schnyder, A.; Spindler, F.; Landolt, H.; Tijani, A. J . Am.
Chem. Soc. 1994, 116, 4062-4066. (b) Breutel, C.; Pregosin, P. S.;
Salzmann, R.; Togni, A. J . Am. Chem. Soc. 1994, 116, 4067-4068. (c)
Togni, A.; Breutel, C.; Soares, M. C.; Zanetti, N.; Gerfin, T.; Gramlich,
V.; Spindler, F.; Rihs, G. Inorg. Chim. Acta 1994, 222, 213-224. (d)
Abbenhuis, H. C. L.; Burckhardt, U.; Gramlich, V.; Ko¨llner, C.;
Pregosin, P. S.; Salzmann, R.; Togni, A. Organometallics 1995, 14,
759-766. (e) Schnyder, A.; Hintermann, L.; Togni, A. Angew. Chem.
1995, 107, 996-998 (Angew. Chem. Int. Ed. Engl. 1995, 34, 931-933).
(f) Burckhardt, U.; Hintermann, L.; Schnyder, A.; Togni, A. Organo-
metallics 1995, 14, 5415-5425. (g) Spencer, J .; Gramlich, V.; Ha¨usel,
R.; Togni, A. Tetrahedron: Asymmetry 1996, 7, 41-44. (h) Zanetti, N.
C.; Spindler, F.; Spencer, J .; Togni, A.; Rihs, G. Organometallics 1996,
15, 860. (i) Togni, A.; Burckhardt, U.; Pregosin, P. S.; Salzmann, R. J .
Am. Chem. Soc. 1996, 118, 1031-1037.
(3) Abbenhuis, H. C. L.; Burckhardt, U.; Gramlich, V.; Togni, A.;
Albinati, A.; Mu¨ller, B. Organometallics 1994, 13, 4481-4493.
0276-7333/96/2315-1614$12.00/0 © 1996 American Chemical Society

Chiral Ferrocenyl and Ruthenocenyl Ligands
Organometallics, Vol. 15, No. 6, 1996 1615
Sch em e 1
Sch em e 2
informative, as it would enable a direct comparison of
structural and conformational aspects of similar Fe(II)
and Ru(II) systems with a particular emphasis on their
respective performances in catalysis.4
This account provides full synthetic details for the
preparation of new enantiomerically enriched rutheno-
cenyl amines, as well as their derivatization to give new
ligands for transition-metal-catalyzed asymmetric reac-
tions. A comparison of the performance of such ligands
with that of their isostructural ferrocenyl analogues
is also presented.
diastereoisomerically pure (S,R)-5 can be easily obtained
by crystallization from CH2Cl2/Et2O mixtures. More-
over, the absolute configuration of 5 was elucidated by
an X-ray structural analysis (vide infra). In the case of
the ruthenocenyl amine 6, which is invariably obtained
as an oil, separation of the two diastereoisomers cannot
be achieved by a simple crystallization procedure nor
by column chromatography. This, however, does not
hamper the obtention of virtually enantiomerically pure
ruthenocenyl derivatives from 6 (vide infra).
Attempts to ortho-lithiate the ruthenocenyl deriva-
tives 5 and 6 using n-BuLi were unsuccessful. As we
recently reported for the related iron chemistry, this
rather inert behavior toward ortho-metalation is as-
cribed to the presence of sterically demanding substit-
uents at the nitrogen center, which may severely
disfavor chelation in the lithiated ruthenocenyl inter-
mediates. With the ruthenocenyl amine 6, this com-
plication can be easily alleviated since virtually quan-
titative conversion into its dimethylamino analogue 7
(see Scheme 1) is readily achieved. In contrast to the
purification of 6, which involves flash chromatography
on silica, attempts to purify 7 by a similar procedure
were unsuccessful and invariably led to decomposition
of the compound to the vinyl complex, Cp*RuC5H4-
CHdCH2. Subsequent lithiation of the ruthenocenyl
amine 7 with n-BuLi occurs with high stereoselectivity,
and the lithiated ruthenocene reacts readily with chlo-
rodiphenylphosphine to give the enantiomerically en-
riched (diphenylphosphino)ruthenocenyl amine (S)-(R)-8
(see Scheme 2) in nearly 70% yield. The diastereo-
selectivity of this reaction must be greater than 98%,
as the (S)-(S)-diastereoisomer could never be identified.
The analogous conversion of the cationic ruthenium
amine 5 into its dimethylamine derivative could not be
achieved. 5 is inert toward Me2NH/HOAc at room
temperature for 48 h or Me2NH‚HCl/KOAc in HOAc at
reflux temperature over similar periods of time. Such
a lack of reactivity in this aminolysis reaction is prob-
ably due to the inability of the metal center to provide
anchimeric assistance prior to solvolysis, as this would
necessitate the generation of an unfavorable dicationic
Resu lts a n d Discu ssion
Ster eoselective Syn th eses of Ru th en ocen yl P ,P
a n d P ,N Liga n d s. The “half-sandwich” complexes
[Cp*Ru(µ3-Cl)]45 and [(p-cymene)RuCl2]26 are convenient
starting materials for the synthesis of the requisite
heteroleptic Ru sandwich compounds. Transmetala-
tions with the cyclopentadienyl synthon 4 proceed
smoothly and afford respectively the cationic ruthenium
amine derivative 5 and the neutral ruthenocenyl amine
6 (see Scheme 1). The latter is obtained in 93% yield
as a dark oil after conventional workup involving its
purification by flash chromatography over silica. The
transmetalation of [(p-cymene)RuCl2]2 with 4, followed
by an anion-exchange reaction with KPF6, furnishes the
cationic mixed sandwich derivative 5 in 72% yield as
an air stable beige/brown solid. Since the cyclopenta-
dienyl synthon (S,R)-4 contains, due to the incomplete
stereoselectivity in its formation (de ) 74%), ca. 13% of
the minor diastereoisomer (R, R)-4, the new heteroleptic
complexes 5 and 6 should exist as approximate 7:1
mixtures of diastereoisomers. The 1H and 13C NMR
resonances of the minor isomers, however, are not
sufficiently dinstinguishable from those of the major
diastereoisomers to allow an assessment of the dia-
stereoisomeric purity of 5 or 6 by NMR. Gratifyingly,
(4) For a recent discussion of the differences between ruthenocenyl
and ferrocenyl ligands. see: Hayashi, T.; Ohno, A.; Lu, S.; Matsumoto,
Y.; Fukuyo, E.; Yanagi, K. J . Am. Chem. Soc. 1994, 116, 4221-4226
and references cited therein.
(5) Fagan, P. J .; Ward, M. D.; Calabrese, J . C. J . Am. Chem. Soc.
1989, 111, 1698-1719.
(6) Bennet, M. A.; Smith, A. J . Chem. Soc., Dalton Trans. 1974, 233.

1616 Organometallics, Vol. 15, No. 6, 1996
Abbenhuis et al.
Ta ble 1. Exp er im en ta l Da ta for th e X-r a y Diffr a ction Stu d y of (S,R)-5, (S*)-(R*)-12, a n d (S*)-(R*)-13
compound
(S,R)-5
(S*)-(R*)-12
(S*)-(R*)-13
formula
mol wt
C26H40F6NPRu
612.6
C45H59F3FeO3P2PdS‚CH2Cl2
C45H59F3O3P2PdRuS
1006.48
0.5 × 0.25 × 0.1
20
monoclinic
P21/c
23.78(7)
9.693(11)
20.83(3)
110.7(2)
4490(15)
4
961.24 + 84.93
cryst dim (mm)
data coll T (°C)
cryst syst
space goup
a (Å)
0.63 × 0.4 × 0.44
0.1 × 0.2 × 0.4
20
20
orthorhombic
P212121
10.575(6)
11.508(7)
23.221(14)
monoclinic
P21/c
21.589(8)
13.825(5)
18.071(7)
105.05(3)
5209(3)
4
b (Å)
c (Å)
R (deg)
V (Å3)
2826(3)
4
1.440
6.66
1264
Z
F(calcd) (g‚cm-3
)
1.424
9.78
2296
1.489
9.05
2064
µ (cm-1
F(000)
)
diffractometer
radiation
measd reflcns
Picker-STOE
STOE Picker
Syntex P21
Mo KR (graphite monochrom), λ ) 0.710 73 Å
0 e h e 10, 0 e k e 11,
0 e h e 20, -13 e k e 0,
0 e h e 14, -7 e k e 7,
0 e / e 22
3.0-4.0.0
ω
-17 e / e 16
3.0-40.0
ω
-15 e / e 14
3.0-40.0
ω
2θ range (deg)
scan type
scan width (deg)
1.00
1.00
1.05
bkgd time (s)
0.30 × scan time
1.0-4.0 in ω
1535
0.25 × scan time
1.0-4.0 in ω
4851
0.25 × scan time
1.0-4.0 in ω
1739
max scan speed (deg‚min-1
no. of indep data coll
no. of obsd reflcns (no)
)
1387
3551
1475
2
2
2
2
2
2
|Fo | > 4.0F(|F| )
|Fo | > 4.0F(|F|
|Fo | > 4.0F(|F| )
face-indexed numerical
0.9079-0.9582
450
abs corr
transm coeff
N/A
N/A
no. of params refined (nv)
quantity minimized
weighting scheme
Ra
304
545
∑w(Fo - Fc)2
∑w(Fo - Fc)2
∑w(Fo - Fc)2
w-1 ) σ2(F) + 0.0010F2
0.0419
w-1 ) σ2(F) + 0.0134F2
w-1 ) σ2(F) + 0.0023F2
0.0595
0.0797
0.75
0.0657
0.0926
1.64
b
Rw
0.0544
1.70
GOFc
a
b
2
R ) ∑(||Fo| - (1/k)|Fc||)/∑|Fo|. Rw ) ∑w(||Fo| - (1/k)|Fc||)2/[∑w|Fo| ]1/2
.
c GOF ) [∑w(|Fo| - (1/k)|Fc|)2/(no - nv)]1/2
.
1
species. Consequently, we aborted further efforts to
synthesize cationic ruthenocenyl phosphines from 5 and
restricted the work presented herein on the use of
(diphenylphosphino)ruthenocenyl amine (S)-(R)-8.
pyrazole ligand (S*)-(R*)-11 by H NMR, the ee of (S)-
(R)-11 was confirmed to be greater than 99%. The
optical purity of the ligands (S)-(R)-9 and 10 was
checked by HPLC (see Experimental Section).
Solid Sta te Str u ctu r es of th e Ru th en ocen yl
Der iva tives (S,R)-5, (S)-(R)-9, a n d (S)-(R)-10. In
order to establish the stereochemical integrity and to
study conformational aspects, we carried out X-ray
crystallographic studies on the new heteroleptic ruthe-
nium sandwich derivatives (S,R)-5, (S)-(R)-9, and (S)-
(R)-10. For comparative purposes it was especially
indispensable to obtain data for the latter two ligands
so as to assess the effect of a change in metal, from Fe
to Ru, on, for example, the relative positions of the aryl
groups of the diphenylphosphine fragment of the upper
Cp ring and the flexibility of the metallocene unit as a
whole. An ORTEP view of the first of this series, (S,R)-
5, is shown in Figure 1, and Table 1 collects crystal data
and refinement parameters. The ruthenium atom is
embedded in a mixed sandwich comprising the η6-bound
p-cymene and η5-bound Cp fragment derived from (S,R)-
4. Bond lengths and angles turn out to be routine and
compare with those of, e.g., the similar cationic mixed
sandwich derivative [Cp*Ru(η6-C6(CH3)6)]+.5 The present
structural study proves the absolute configuration S at
the C(6) stereogenic center, thus directly establishing
the stereochemistry of the preferred diastereoisomer of
cyclopentadienyl 4.
As with its iron analogue, the amine (S)-(R)-8 can be
easily obtained in highly enantiomerically enriched form
(>99% ee) by selective crystallization of the racemate
(S*)-(R*)-8 from ethanol solutions of the enantiomeri-
cally enriched compound. It is important to note that
the multistep synthesis of 8 from the chiral Cp synthon
4 is accompanied with no net loss of optical activity. This
can be derived from analysis of the relative amounts of
(S*)-(R*)-8 and (S)-(R)-8 which nicely mirror the dia-
stereoselectivity observed during the generation of 4.
Subsequent derivatization of (S)-(R)-8 by applying a
series of standard transformations, all of which have
precedence in the related ferrocenyl series,3 enabled
access to ligands (S)-(R)-9-11 (see Scheme 2). The first
member of this series, (S)-(R)-9, was obtained as yellow
crystals in 62% yield, after chromatographic purifica-
tion, by the substitution of the dimethylamino group of
(S)-(R)-8 with dicyclohexylphosphine in acetic acid. The
phobyl (9-phospha[3.3.1]bicyclonon-9-yl) incorporating
compound (S)-(R)-10 was equally prepared from (S)-
(R)-8 as a yellow solid in 40% yield in an analogous
manner using a technical mixture of phobane as phos-
phine source.2d The yellow solid (S)-(R)-11 was synthe-
sized from (S)-(R)-8 using 3,5-dimethylpyrazole as
nucleophile. Given that the diamagnetic shift reagent
R-(-)-1-(9-anthryl)-2,2,2-trifluoroethanol was effective
for the separation of the resonances of the racemic
Crystal data and ORTEP views of compounds 9 and
10 are provided as Supporting Information. The struc-
ture of derivative 9 can be compared to that of the

Chiral Ferrocenyl and Ruthenocenyl Ligands
Organometallics, Vol. 15, No. 6, 1996 1617
F igu r e 1. ORTEP view of the cation (S,R)-5 (30% prob-
ability ellipsoids). The Ru atom is located at 1.831 Å from
the Cp ring and 1.703 Å from the p-cymene ring. The
planes of the two rings subtend an angle of 3.8°.
F igu r e 3. ORTEP view of the cation (S*)-(R*)-12 (30%
probability ellipsoids).
Str u ctu r a l Com p a r ison of P a lla d iu m η3-Allyl
Com p lexes In cor p or a tin g Ru th en ocen yl a n d F er -
r ocen yl P h osp h in es. Having established a marked
similarity of the structural/conformational attributes of
the Ru-containing ligands, when compared to the cor-
responding Fe derivatives, it was now necessary to
compare the two systems as ligands. We opted for the
facile synthesis and structural characterization of their
cationic palladium η3-allyl complexes, following proce-
dures already employed in our laboratory.2d The race-
mic ruthenocene (S*)-(R*)-9 and its isostructural bis-
(phosphine) iron analogue (S*)-(R*)-14 were chosen as
ligands (eq 1).
F igu r e 2. Schematic superposition of the structures of (S)-
(R)-10 and (S)-(R)-15,3 showing the virtually identical
conformation in the solid state.
corresponding Fe parent compound, J osiphos, 1, re-
ported recently from these laboratories.2c The only
significant difference, besides the Ru-Cp (vs Fe-Cp)
distances, pertains to the relative orientation of the
cyclohexyl rings, indicating that the PCy2 group con-
stitutes the conformationally most flexible part of the
molecule. The structural features of compound 10 are,
not surprisingly, virtually identical to those of the
corresponding Fe derivative 15 (see Chart 2), discussed
in detail previously.3 The only obvious difference is the
increased Cp-Cp* distance in 10 (Ru-Cp 1.824 Å and
Ru-Cp* 1.808 Å), as compared to 15. The similarity
of derivative 10 with its Fe congener is illustrated in
the schematic superposition of the two structures given
as Figure 2.
The single-crystal X-ray structures of the resulting
palladium η3-allyl complexes (S*)-(R*)-12 and (S*)-(R*)-
13 were consequently determined. Their ORTEP views
are shown in Figures 3 and 4, respectively. Crystal and
refinement parameters are given in Table 1, and Table
2 collects pertinent bond lengths and angles, and torsion
angles for both compounds.
Again, for the sake of simplicity, it is appropriate to
compare the structure of complex 12 with that of the
known compound containing the nonmethylated ligand
J osiphos (1).2c The superposition of the two structures,
shown in Figure 5, indicates a strong similarity of the
conformational features. At first sight, the only relevant
effect due to the addition of five 1′-5′ methyl groups in
12 is shown by the relative position and orientation of
the PPh2 group. It turns out that the latter is “pushed
up” even more than in the parent compound. This is
On the basis of the conformational characteristics
observed for the free ligands 9 and 10, as compared with
similar Fe derivatives, one would anticipate no impor-
tant differences between the Ru- and Fe-containing
ligands, respectively, in their coordination behavior and
hence in the catalytic performances of their complexes.
As we will discuss below, this conclusion turns out to
be incorrect, because of further unexpected features of
the complexes containing the heavier element Ru.

1618 Organometallics, Vol. 15, No. 6, 1996
Abbenhuis et al.
F igu r e 5. Schematic superposition of the structures of the
cation (S*)-(R“)-12 and the corresponding complex contain-
ing the ligand J osiphos (1),2c showing the very similar
conformation in the solid state.
matches very well the one of the free parent ligand
J osiphos. On the basis of these considerations only, one
is tempted to assume that the replacement of the
unsubstituted Cp by Cp*, and, by extension, even more
so the introduction of Ru instead of Fe, would not have
very significant consequences on the conformational
properties of the respective complexes. Therefore, if
ground-state conformational considerations on com-
plexes are important in the discussion of the catalytic
performance of the different ligands, it could be antici-
pated that the three ligands J osiphos, 9, and 14 should
behave in a comparable manner.
F igu r e 4. ORTEP view of the cation (S*)-(R*)-13 (30%
probability ellipsoids).
Ta ble 2. Selected Bon d Dista n ces (Å),a An gles
(d eg),a a n d Tor sion An gles (d eg) for (S*)-(R*)-12
a n d (S*)-(R*)-13
(S*)-(R*)-12
(R*)-(S*)-13
Bond Distances
Pd-P(1)
Pd-P(2)
Pd-C(37)
Pd-C(38)
Pd-C(39)
M-Cp*
2.295(3)
2.312(3)
2.207(15)
2.129(20)
2.180(13)
1.678
2.285(8)
2.353(8)
2.163(27)
2.108(35)
2.118(25)
1.805
The structure of the Ru derivative 13, however, shows
some unusual and unexpected features. First of all, the
orientation of the Pd-allyl fragment turns out to be
completely different from that found in the analogous
complex 12. The Pd atom seems to adopt the most
remote position from the ruthenocene core. This is
illustrated by the large angle subtended by the planes
of the Cp ring and the one defined by the atoms Pd,
P(1), and P(2) of 76° (vs 22.5° for the corresponding
angle in 12). The most astonishing related feature is
the orientation of the two Ph groups on P(1). Both are
in a pseudo-equatorial position, pointing “down”, i.e.
toward the Cp*Ru fragment. The steric repulsion
between the Ph groups and the Cp* ligand (shortest
nonbonding distance is 3.47 Å for C(27)-C(32)) is
relieved by the severe distortion around C(1), with P(1)
located 0.56 Å above the Cp plane. This distortion is
in part also transmitted to C(2), with C(6) positioned
at 0.25 Å (vs 0.01 Å in 12) from the same plane. These
features seem to indicate an enhanced deformability of
the ruthenocene core, as compared to ferrocene, in this
class of compounds. A tentative conclusion from these
observations is that complexes containing ruthenocenyl
ligands will be conformationally more flexible than their
ferrocenyl counterparts.
M-Cp
1.661
1.819
Bond Angles
P(1)-Pd-P(2)
P(1)-Pd-C(37)
P(2)-Pd-C(39)
Pd-P(1)-C(1)
Pd-P(2)-C(6)
C(37)-C(38)-C(39)
Cp-Cp*
95.3(1)
95.2(4)
102.8(4)
116.6(4)
106.3(3)
141.4(23)
9.0
94.0(2)
100.3(6)
99.3(7)
106.2(6)
112.4(6)
148.7(36)
12.6
Torsion Angles
Pd-P(1)-C(1)-C(2)
Pd-P(2)-C(6)-C(2)
P(1)-C(1)-C(5)-C(4)
C(4)-C(3)-C(2)-C(6)
P(2)-C(6)-C(2)-C(1)
8
73
162
180
62
53
48
158
167
68
a
Numbers in parentheses are esd’s in the least significant
digits.
clearly because of severe nonbonding interactions be-
tween the protruding Cp* ligand and the C(26)-C(31)
phenyl group, as illustrated by (1) the short distance
between the plane defined by the latter and the methyl
group C(32) of 3.33 Å (shortest atom to atom separation
is C(28)-C(32) of 3.53 Å), (2) the Cp-Cp* angle of 9°,
and (3) the out-of-plane position of the phosphorus atom
P(1). Its distance from the plane of the “upper” Cp ring
is now 0.49 Å (vs 0.29 Å for the parent compound).
Whereas in the J osiphos-containing complex one could
describe the two phenyl groups in terms of axial and
equatorial positions for the “lower” and the “upper” aryl,
respectively, this distinction no longer applies to 12.
Both substituents on P(1) assume pseudo-equatorial
positions. The conformation of the rest of the molecule
is very similar in both compounds. Finally, from a
qualitative point of view, the overall conformation of the
Fe-containing ligand 14 in the Pd-allyl complex 12
Asym m etr ic Ca ta lysis In volvin g th e Ru th en o-
cen yl Der iva tives (S)-(R)-9-11 a s Liga n d s. Carry-
ing on from our very successful exploitation of ligands
1-3 in a wide range of asymmetric processes, often
giving rise to ee’s well in excess of 90%, a few catalytic
applications of the new ruthenocenes (S)-(R)-9-11 were
attempted. Two routine reactions were performed, viz.
the palladium-catalyzed alkylation of 1,3-diphenyl-3-
acetoxypropene with dimethyl malonate (see Table 3),7
and the rhodium-catalyzed hydroboration of styrene (see
Table 4).8 As a comparison, the recently introduced

Chiral Ferrocenyl and Ruthenocenyl Ligands
Organometallics, Vol. 15, No. 6, 1996 1619
Ta ble 3. P a lla d iu m -Ca ta lyzed Alk yla tion of
1,3-Dip h en yl-3-Acetoxyp r op en e w ith Dim eth yl
Ma lon a tea
employed low conversions were observed even after as
much as 92 h. The enantioselectivities observed for
these reactions are equally rather low, reaching a
maximum of 78%, after 0.75 h of reaction, for the
ferrocenyl ligand (S)-(R)-14, although the activity of this
catalyst is hampered by its low stability and rapid
deactivation; even after 50 h only 28% conversion is
attained. It is pertinent to note that the nonmethylated
parent ligand 1 gave 93% ee and 99% yield for the same
reaction after only 3 h. For the ruthenocenyl ligands
employed in these tests very disappointing ee’s and
activities were observed, (S)-(R)-10 and (S)-(R)-11 giving
ee’s much lower than 30%, and the opposite enantiomer
was obtained in each case. As with the ferrocene (S)-
(R)-14, these reactions (entries 5-8) suffered from
catalyst deactivation over prolonged periods.
entry
ligand
time (h) conversion (%) % ee (abs conf)
1
2
3
4
5
6
7
8
(S)-(R)-9
(S)-(R)-9
0.5
36
0.75
50
44
92
44
96
53
100
15
28
22
26
10
100
77 (R)
73 (R)
78 (R)
77 (R)
21 (S)
18 (S)
28 (S)
27 (S)
(S)-(R)-14
(S)-(R)-14
(S)-(R)-10
(S)-(R)-10
(S)-(R)-11
(S)-(R)-11
a
Catalytic experiments were carried out as described in ref 2a
using 1.0 mol % of catalyst at 20 °C.
Ta ble 4. Rh od iu m -Ca ta lyzed Hyd r obor a tion of
Styr en ea
regio-
The rhodium-catalyzed hydroboration of styrene con-
tinues to be an active subject of research in ours and
other laboratories.2e,8 Our results obtained with the new
ligands presented in Table 4 show certain useful trends.
Once again, the incorporation of a Cp* fragment into
the ferrocenyl ligand (S)-(R)-14 has a detrimental effect
on the enantioselectivity. Thus, whereas (R)-(S)-1 is an
extremely useful ligand for this reaction (91.5% ee at
-78°C), (S)-(R)-14 is rather disappointing (54% ee, entry
3, Table 4), as is the corresponding ruthenocene, (S)-
(R)-9 (40% ee). In line with our previously reported
observations, the best results are obtained with ligands
incorporating pyrazole functions, although in these
cases the regioselectivity of the reaction is rather low.
Thus, the ruthenocenyl derivative (S)-(R)-11 as ligand
yields (S)-1-phenylethanol with an ee of 87% and 67%
regioselectivity, whereas the ferrocenyl derivative (S)-
(R)-16 yields the same alcohol with an ee of 94%
accompanied with a significantly higher regioselectivity
(76%) (entries 7 and 8). The highest regio and enantio-
selectivities were attained with the latter ligand at 20
°C.
entry
ligand
T (°C) time (h) selectivity % ee (abs conf)
1
2
3
4
5
6
7
8
9
(S)-(R)-9
(S)-(R)-9
20
0
20
20
0
20
20
20
0
7
7
18
7
95
99
95
89
88
77
67
76
68
40 (S)
37 (S)
54 (S)
34 (S)
26 (S)
32 (S)
87 (R)
94 (R)
92 (R)
(S)-(R)-14
(S)-(R)-10
(S)-(R)-10
(S)-(R)-15
(S)-(R)-11
(S)-(R)-16
(S)-(R)-16
7
18
18
18
18
a
Catalytic experiments were carried out as described in ref 2a
using 1.0 mol % of catalyst. Complete conversion of the starting
material styrene was achieved in all cases listed here. Regioselec-
tivity refers to % of 1-phenylethanol.
Ch a r t 2
Con clu sion s
Hayashi et al. recently reported an improvement in
the enantioselectivity of a number of palladium-cata-
lyzed allylic substitution reactions on moving from a
ferrocenyl to a ruthenocenyl bis(phosphine) system,
whereby one phosphino fragment was situated on the
upper and the other on the lower Cp ring.4 Here, the
greater distance between the cyclopentadienyl rings in
the ruthenocene system leads to a larger bite angle. This
was postulated to engender a tighter arrangement of
the substituents on the phosphorus atoms, thus creating
a more rigid “chiral pocket”. Our system is significantly
different from the latter in that in our case the two
chelating fragments, viz. PP or PN functions, are located
on the same, upper cyclopentadienyl ring. The unex-
pected structural features of complex 13 are interpreted
as an expression of the greater conformational flexibility
associated with our ruthenocenyl system, translating
into a less efficient transmission of the chiral informa-
tion in the catalytic reactions. In the Cp* ferrocene
system, the larger lower ring seems to be responsible
for a general decrease in activity in the Pd-catalyzed
allylic chemistry, probably because of a much slower
oxidative addition of the substrate, caused by unfavor-
able steric interactions. On the other hand, the Cp*
ligand does not have any detrimental effects in the Rh-
isostructural ferrocenyl derivatives (S)-(R)-14-16 were
also tested (see Chart 2).
From Table 3, a few important conclusions can be
drawn. It is apparent that the activities of the catalysts
are rather low; quantitative conversion of the allylic
substrate when employing compounds (S)-(R)-9 and (S)-
(R)-11 is only achieved after 36 and 96 h, respectively
(entries 2 and 8, Table 3). For the other catalysts
(7) For reviews, see: (a) Consiglio, G.; Waymouth, R. M. Chem. Rev.
1989, 89, 257-276. (b) Hayashi, T. In Catalytic Asymmetric Synthesis;
Ojima, I., Ed.; VCH: New York, 1993; pp 325-365. (c) Godleski, S. A.
In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, U.K., 1991; Vol. 4, Chapter 3.3, pp 585-661. (d)
Frost, C. G.; Howarth, J .; Williams, J . M. J . Tetrahedron: Asymmetry
1992, 3, 1089-1122. For a recent example, see: (e) von Matt, P.; Lloyd-
J ones, G. C.; Minidis, A. B. E.; Pfaltz, A.; Macko, L.; Neuburger, M.;
Zehnder, M.; Ru¨egger, H.; Pregosin, P. S. Helv. Chim. Acta 1995, 78,
265-284 and references cited therein.
(8) For a recent review, see: (a) Burgess, K.; Ohlmeyer, M. J . in
Homogeneous Transition Metal Catalyzed Reactions (Adv. Chem. Ser.
230); (Moser, W. R., Slocum, D. W., Eds.; American Chemical Society:
Washington, DC, 1992; pp 163-177 and references cited therein. For
specific examples, see: (b) Hayashi, T.; Matsumoto, Y.; Ito, Y. J . Am.
Chem. Soc. 1989, 111, 3426-3428. (c) Matsumoto, Y.; Hayashi, T.
Tetrahedron Lett. 1991, 32, 3387-3390. (d) Hayashi, T.; Matsumoto,
Y.; Ito, Y. Tetrahedron: Asymmetry 1991, 2, 601-612. (e) Burgess, K.;
van der Donk, W. A.; Ohlmeyer, M. J . Tetrahedron: Asymmetry 1991,
2, 613-621. (f) Brown, J . M.; Hulmes, D. I.; Layzell, T. P. J . Chem.
Soc., Chem. Commun.1993, 1673-1674.

1620 Organometallics, Vol. 15, No. 6, 1996
Abbenhuis et al.
added and the water layer extracted with hexane (3 × 200
mL). The combined extracts were dried over MgSO4 followed
by removal of the solvent in vacuo. The residual brown oil
was purified by chromatography on silica using hexane
(containing ca. 5% of Et3N). Yield: 8.0 g (93%) of dark oil. 1H
NMR (250 MHz, CDCl3): δ 4.1 (m, 4H, C5H4), 3.35 and 2.38
(both q, 2 × 1H, CH(Me)N), 2.1 (s, 3H, NMe), 1.9 (s, 15H,
C5Me5), 1.17 and 0.75 (both d, 2 × 3H, CHMeN). 13C NMR
(63 MHz, CDCl3): δ 93.4 (ipso-C of C5H4), 84.4 (ipso-C of
C5Me5), 73.4, 72.6, 71.8, 70.9 (C5H4), 58.4 (CpCH), 56.5
(CpCHMeNMeCH), 42.2 (NMe), 31.1-26.7 (Cy), 15.4 and 13.4
(both CHMeN), 11.8 (C5Me5). Anal. Calcd for C26H41NRu: C,
66.63; H, 8.82; N, 2.99. Found: C, 66.79; H, 8.63; N, 2.88.
(S)-Cp *Ru C5H4CH(Me)NMe2 ((S)-7). To a cooled (ca. -10
°C) dark solution of (S,R)-6 (13.8 g, 29.3 mmol) in HNMe2 (25
mL) was slowly added 45 mL of AcOH. The resulting solid
was heated for 30 min at 60 °C, giving a clear dark solution.
Water (ca. 100 mL) was added and the pH adjusted to ca. 10
by careful addition of NaOH. Extraction with hexane (3 × 200
mL), followed by drying of the combined extracts over MgSO4
and removal of the solvent in vacuo, gave a dark oil that
contained the product together with HN(Me)CH(Me)Cy. This
sec-amine was distilled off (100 °C, 0.1 Torr) leaving the
product as a brown oil; yield 10.2 g (93%). 1H NMR (250 MHz,
CDCl3): δ 4.1 (m, 4H, C5H4), 3.24 (q, 1H, CHMe), 2.1 (s, 6H,
NMe2), 1.9 (s, 15H, C5Me5), 1.23 (d, 3H, CHMe). 13C NMR (63
MHz, CDCl3): δ 89.5 (ipso-C of C5H4), 84.1 (ipso-C of C5Me5),
73.2, 72.2, 71.8, 70.4 (C5H4), 57.5 (CpCH), 40.4 (NMe2), 14.6
(CHMeN), 11.6 (C5Me5). Anal. Calcd for C19H29NRu: C, 61.26;
H, 7.85: N, 3.76. Found: C, 61.74; H, 7.68; N, 3.40.
catalyzed hydroboration of styrene, and indeed ligand
16 gives virtually identical results as its nonmethylated
congener.1e We interpret this result as indicative of very
similar structural/conformational properties of the two
catalytically active complexes, not being influenced by
the steric nature of the lower Cp ring.
The present work has shown that structural varia-
tions in peripheral regions of metallocene ligands may
change their catalytic properties in a drastic manner.
We demonstrated that (1) the introduction of a Cp*
fragment in our ferrocenyl ligands, instead of a non-
substituted cyclopentadienyl, and (2) the replacement
of Fe by Ru, respectively, has different consequences on
their catalytic performances. For diphosphine deriva-
tives forming six-membered chelate rings, both changes
have detrimental effects. In the case of the pyrazole-
containing derivatives forming seven-membered che-
lates, the influence of both the Cp* and ruthenium is
much less pronounced.
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All reactions with air- or mois-
ture-sensitive materials were carried out under Ar using
standard Schlenk techniques. Freshly distilled, dry, and
oxygen-free solvents were used throughout. Technical grade
phobane was obtained by courtesy of Prof. A. Salzer, RWTH
Aachen, and was used as received. Routine 1H (250.133 MHz),
13C (62.90 MHz), and 31P (101.26 MHz) spectra were recorded
with a Bruker AC 250 spectrometer. Chemical shifts are given
in ppm and coupling constants (J ) are given in hertz. Merck
silica gel 60 (70-230 mesh) was used for column chromatog-
raphy. Optical rotations were measured with a Perkin-Elmer
341 polarimeter using 10 cm cells. Elemental analysis were
performed by the “Mikroelementar-analytisches Laboratorium
der ETH”. Catalytic experiments and analysis of reaction
products were carried out as previously described.2a
(S)-(R)-Cp *Ru C5H3CH(Me)NMe2P P h 2-2 ((S)-(R)-8).
A
solution of (S)-7 (10.2 g, 27.4 mmol) and BuLi (19 mL 1.6 M
(30.4 mmol in hexane) in Et2O (50 mL) was stirred for 2 days.
Subsequently, PPh2Cl (6.0 mL, 32 mmol) was added and the
resulting brown suspension was refluxed for 4 h followed by
careful addition of saturated NaHCO3 (ca. 30 mL). Extraction
with toluene (3 × 100 mL) followed by drying of the combined
organic layers on MgSO4 and removal of the solvent in vacuo
gave a reddish oil. This was subjected to flash chromatography
over silica using hexane (containing ca. 5% of Et3N) in order
to elute impurities followed by elution of the product with THF.
Finally, the product was purified by flash chromatography over
Al2O3 using toluene (containing ca. 5% of Et3N) affording 10.2
g (67%) of product. Crystallization from ethanol gave 1.7 g of
(S,R)-[(p-cym en e)Ru (C5H4CH(Me)N(Me)CH(Me)Cy)]+-
[P F 6]- ((S,R)-5). A cooled THF solution (-40 °C) of LiC5H4-
CH(Me)N(Me)CH(Me)Cy (freshly prepared from 0.82 g (3.8
mmol) of (R)-C5H4dCHN(Me)CH(Me)Cy and 2.4 mL of 1.6 M
MeLi (3.8 mmol) at 0 °C) was transferred via cannula to a THF
(15 mL) suspension of [(p-cymene)RuCl2]2 (1.1 g, 1.9 mmol)
that was also kept at -40 °C. After being warmed up to room
temperature and stirred for 1.5 h, the resulting red solution
was filtered and the filtrate evaporated in vacuo. To the
resulting red oil was added MeOH (50 mL) and excess KPF6
(1.3 g, 7.1 mmol). After 1 h of stirring at room temperature,
evaporation of the solvent afforded a brown solid that was
subsequently extracted with CH2Cl2 (3 × 70 mL), filtered to
remove impurities and excess inorganic salts, and concentrated
in vacuo to ca. 10 mL. Addition of Et2O (60 mL) induced
20
crystalline, pale yellow, almost racemic product ([R]D ) 14
(CHCl3, c ) 1.0)) while removal of the solvent from the mother
liquor in vacuo left 8.5 g of almost optically pure compound
20
as a red oil ([R]D ) 248 (CHCl3, c ) 0.44)). 1H NMR (250
MHz, CDCl3): δ 7.62 (m, 2H, PPh2), 7.27 (m, 8H, PPh2), 4.23
3
4
(m, 3H, C5H3), 3.65 (dq, 1H, CHMeN, J (H,H) ) 6.5, J (P,H)
) 1.7), 1.90 (s, 15H, C5Me5), 1.78 (s, 6H, NMe2), 0.98 (d, 3H,
CH(Me)N). 13C NMR (63 MHz, CDCl3): δ 135.0-126.8 (non
quartenary C of PPh2), 85.1 (C5Me5), 75.9, 74.8, 74.0 (non
quartenary C of C5H3), 56.1 (CHMeN), 38.4 (NMe2), 11.5
(C5Me5), 7.2 (CHMeN). 31P NMR (101 MHz, CDCl3): δ -24.6
(s, PPh2). Anal. Calcd for C31H38NPRu: C, 66.88; H, 6.88:
N, 2.52. Found: C, 66.95; H, 6.88; N, 2.47.
20
precipitation of 1.7 g (72%) of brown product ([R]D ) 7.6
(CH2Cl2, c ) 0.1)). 1H NMR (250 MHz, CDCl3): δ 6.01 (m,
4H, Ar), 5.32 (s, 1H, C5H4), 5.22 (m, 3H, C5H4), 3.51 (q, 1H,
3J (H,H) ) 6.8, CHMe), 2.68 (q, 1H, CHMe2, 3J (H,H) ) 6.8),
2.32 (m, 3H, Me), 2.02 (s, 3H, Me), 1.69 (m, 3H, NMe), 1.24
(S)-(R)-Cp *Ru C5H3CH(Me)P Cy2P P h 2-2 ((S)-(R)-9).
A
solution of (S)-(R)-8 (0.95 g, 1.7 mmol) and HPCy2 (0.38 mL,
1.9 mmol) in AcOH (35 mL) was stirred at 80 °C for 2 h. The
solvent was removed in vacuo and the sticky residue subjected
to flash chromatography on Al2O3 using hexane/toluene (3:1,
containing 5% Et3N) as eluent. The product was obtained
analytically pure after crystallization from a minimum of hot
3
(d, 6H, Me2CH, J (H,H) ) 6.8), 1.91-0.75 (m, 11H, Cy). 13C
NMR (63 MHz, CDCl3): δ 136.0, 117.6, 87.1, 86.8, 84.5, 84.4,
80.7 (C5H4 and Ar), 59.8, 55.9 (CHMeN), 42.0 (NMe), 32.0, 31.0,
30.7, 29.9, 26.6, 26.5 (Cy), 23.4 (CHMe2), 19.7 (CHMe2), 16.4,
13.5 (CHMeN). MS (FAB) (m/e) 468 (M+, 100%). Anal. Calcd
20
for
C26H40NF6PRu‚CH2Cl2: C, 46.49; H, 6.07; N, 2.01.
EtOH; yield 0.75 g (62%) of pale yellow crystals ([R]D ) 253
Found: C, 46.74; H, 6.03; N, 2.00.
(CHCl3, c ) 1.1)). Note: the racemic compound (prepared via
the same method from racemic Cp*RuC5H3CH(Me)NMe2PPh2-
2) can be separated on a Daicel Chiracel OD-H column
(hexane/2-propanol ) 99.5:0.5, flow ) 1.0 mL‚min-1) with Rt
(R)-(S) ) 3.34 min and Rt (S)-(R) ) 3.44 min. 1H NMR (250
MHz, CDCl3): δ 7.63 (m, 2H, PPh2), 7.25 (m, 8H, PPh2), 4.44,
4.13 (m, 3H, C5H3), 2.79 (dq, 1H, CHMeP, 3J (H,H) ) 5.4,
(S,R)-Cp *Ru C5H4CH(Me)N(Me)CH(Me)Cy ((S,R)-6). To
a solution of LiC5H4CH(Me)N(Me)CH(Me)Cy in ca. 30 mL of
THF (prepared from 4.4 g (20 mmol) of (R)-C5H4dCHN(Me)-
CH(Me)Cy and 13 mL of 1.6 M MeLi (21 mmol) at 0 °C) was
added [Cp*Ru(µ3-Cl)]4 (5.0 g, 4.6 mmol). The resulting dark-
brown solution was stirred for 30 min. Water (150 mL) was

Chiral Ferrocenyl and Ruthenocenyl Ligands
Organometallics, Vol. 15, No. 6, 1996 1621
J (P,H) ) 2.2 Hz), 1.73 (s, 15H, C5Me5), 1.38 (dd, 3H, CHMeP,
3J (H,H) ) 5.7, 3J (H,P) ) 7.1), 1.8-1.3 (m, 22H, Cy). 13C NMR
(63 MHz, CDCl3): δ 135.7-126.7 (nonquartenary C of PPh2),
85.0 (C5Me5), 75.6, 74.5, 74.1 (nonquartenary C of C5H5), 32.7-
24.7 (PCy2), 18.2 (CHMeP), 11.6 (C5Me5). 31P NMR (101 MHz,
CDCl3): δ -26.3 (d, PPh2, 4J (P,P) ) 36 Hz), 11.5 (d, PCy2,
4J (P,P) ) 36 Hz). Anal. Calcd for C41H54P2Ru: C, 69.37; H,
7.67. Found: C, 69.62; H, 7.62.
of [PdCl(η3-C3H5)]2 (52 mg, 0.14 mmol) and (S*)-(R*)-
Cp*FeC5H3(CH(Me)PCy2)(PPh2)-1,2 (199 mg, 0.30 mmol) in
CH2Cl2 (10 mL) was added Ag[CF3SO3] (77 mg, 0.30 mmol),
dissolved in 1 mL of MeOH. After 1 h, the resulting brown
suspension was filtered over Celite and the filtrate evaporated
in vacuo. The orange spongy residue was dissolved in 10 mL
of CH2Cl2 and layered with 50 mL of hexane. From this
system, 210 mg (70%) of crystalline red product could be
obtained overnight. 31P NMR (101 MHz, CDCl3): 2 isomers
in approximate 1:2 ratio; minor isomer, δ 61.1 (d, PPh2, 4J (P,P)
) 49 Hz), 17.0 (d, PCy2, 4J (P,P) ) 49 Hz); major isomer, δ 58.7
(d, PPh2, 4J (P,P) ) 46 Hz), 15.0 (d, PCy2, 4J (P,P) ) 46 Hz).
(S)-(R)-Cp *Ru C5H3CH(Me)P C8H14P P h 2-2 ((S)-(R)-10).
The procedure is the same as that for (S)-(R)-9 except that
(S)-(R)-8 (0.40 g, 0.72 mmol) and HPC8H14 (1.28 g, 8.8 mmol,
2:1 mixture of [3.3.1] and [4.2.1] isomers) reacted to give 0.19
20
1
g (40%) of pale yellow crystalline product ([R]D ) +182,
The H and 13C NMR spectra were poorly resolved indicating
(CHCl3, c ) 0.70)). Note: the racemic compound (prepared
via the same method from racemic Cp*RuC5H3CH(Me)NMe2-
PPh2-2) can be separated on a Daicel Chiracel OD-H column
(hexane/2-propanol ) 99.0:1.0, flow ) 0.5 mL‚min-1) with Rt
(R)-(S) ) 7.19 min and Rt (S)-(R) ) 7.44 min. 1H NMR (250
MHz, CDCl3): δ 7.58 (m, 2H, PPh2), 7.26 (m, 8H, PPh2), 4.65,
fluxionality in solution. Anal. Calcd for C45H59O3F3P2FePdS‚
1/2CH2Cl2: C, 54.48; H, 6.03. Found: C, 54.24; H, 5.91.
(S*)-(R*)-[(Cp *Ru C5H3(CH(Me)P Cy2)(P P h 2)-1,2)P d (η3-
C3H5)]+[OTf]- ((S*)-(R*)-13). The procedure is the same as
that for (S*)-(R*)-12 except that [PdCl(η3-C3H5)]2 (24 mg, 0.07
mmol), (S*)-(R*)-9 (100 mg, 0.14 mmol) in CH2Cl2 (5 mL), and
Ag[CF3SO3] (36 mg, 0.14 mmol) reacted to give ca. 70 mg (50%)
of crystalline product. 31P NMR (101 MHz, CDCl3): 2 isomers
in approximate 2:3 ratio; minor isomer, δ 64.9 (d, PPh2, 4J (P,P)
) 47 Hz), 14.4 (d, PCy2, 4J (P,P) ) 47 Hz); major isomer, δ 67.8
(d, PPh2, 4J (P,P) ) 47 Hz), 15.8 (d, PCy2, 4J (P,P) ) 47 Hz).
3
4.10, and 4.06 (m, 3H, C5H3), 2.80 (q, 1H, CHMeP, J (H,H) )
6.2), 1.76 (s, 15H, C5Me5), 2.0-1.0 (m, 14H, C8H14), 1.40 (dd,
3H, CHMeP, 3J (H,H) ) 6.2, 3J (H,P) ) 13). 13C NMR (63 MHz,
CDCl3): δ 140.8-127.8 (nonquartenary C of PPh2), 85.7
(C5Me5), 75.2, 75.4, 73.8 (non quartenary C of C5H3), 32.4-
23.3 (PC8H14), 21.5 (CHMeP), 12.2 (C5Me5). 31P NMR (101
MHz, CDCl3): δ -25.6 (s, PPh2), -17.8 (s, PCy2). Anal. Calcd
for C37H46P2Ru: C, 67.97; H, 7.09. Found: C, 67.92; H, 6.83.
Ra cem ic (S*)-(R*)-Cp *Ru C5H3CH(Me){N2C3HMe2-3,5)-
P P h 2}-1,2 ((S*)-(R*)-11). A pale yellow solution of (S*)-(R*)-8
(0.60 g, 1.1 mmol) and 3,5-dimethylpyrazole (1.60 g, 16.6 mmol)
in 5 mL of AcOH was heated at 60 °C for 30 min. Water (60
mL) was added, and the resulting suspension was made basic
by careful addition of NaOH, followed by extraction with
hexane (100 mL). The hexane extract was dried over MgSO4
and the solvent removed in vacuo, leaving a crude product that
was crystallized from ca. 50 mL of hot EtOH/H2O (4:1) to give
0.32 g (49%) of pale yellow product.
Anal. Calcd for
C45H59O3F3P2RuPdS: C, 53.70; H, 5.91.
Found: C, 53.83; H, 6.09.
X-Ra y Cr ysta llogr a p h ic Stu d ies of 5, 9, 10, 12, a n d 13.
Selected crystallographic and relevant data collection param-
eters for 9, 12, and 13 are listed in Table 1. Data were
measured with variable scan speed to ensure constant statisti-
cal precision on the collected intensities. One standard
reflection was measured every 120 reflections; no significant
variation was detected.
The structures were solved either by direct (5, 10, 12, 13)
or Patterson (9) methods and refined by full-matrix least-
squares using anisotropic displacement parameters for all non-
hydrogen atoms. The contribution of the hydrogen atoms in
their idealized position (riding model with fixed isotropic U )
0.080 Å2) was taken into account but not refined. All calcula-
tions were carried out by using the Siemens SHELXTL PLUS
system.
(S)-(R)-Cp*Ru C5H3CH(Me){N2C3HMe2-3,5)P P h 2}-1,2 ((S)-
(R)-11). The procedure is the same as above, except that (S)-
(R)-8 (0.65 g, 1.2 mmol) and 3,5-dimethylpyrazole (0.90 g, 9.3
mmol) reacted to give 0.80 g of crude product, contaminated
by excess dimethylpyrazole. The compound could be obtained
analytically pure in ca. 10% yield by repeated crystallization
from EtOH/H2O. 1H NMR (250 MHz, CDCl3): δ 7.51 (m, 2H,
PPh2), 7.25 (m, 3H, PPh2), 6.98 (q, 3H, PPh2), 6.85 (t, 2H,
PPh2), 5.14 (dq, 1H, CHMeN, 3J (H,H) ) 6.8, 4J (P,H) ) 1.9),
4.99 (s, 1H, N2C3H), 4.73, 4.30, and 4.01 (m, 3H, C5H3), 2.12,
1.93 (s, 2 × 3H, N2C3HMe2), 1.81 (s, 15H, C5Me5), 1.59 (d, 3H,
CHMeN, 3J (H,H) ) 6.9). 13C NMR (63 MHz, CDCl3): δ 135.5-
126.8 (nonquartenary C of PPh2), 104.2 (p-C of C3N2), 85.5
(C5Me5), 76.0, 75.4 (nonquartenary C of C5H3), 51.3 (d,
Ack n ow led gm en t. This research was in part sup-
ported by the Swiss National Science Foundation,
program CHiral2 (postdoctoral grants to H.C.L.A. and
J .S.).
Su p p or tin g In for m a tion Ava ila ble: Tables of crystal
data and refinement details, complete atomic coordinates and
U values, complete bond distances and angles, and anisotropic
displacement coefficients for the non-carbon atoms for com-
pounds 5, 9, 10, 12, and 13 and ORTEP views and atom-
numbering schemes for compounds 9 and 10 (Figures S1 and
S2) (52 pages). Ordering information is given on any current
masthead page.
3
CHMeN, J (P,H) ) 7), 20.0, 13.7 (N2C3HMe2), 11.4 (CHMeP,
4J (P,H) ) 9), 11.7 (C5Me5). 31P NMR (101 MHz, CDCl3): δ
-26.5 (s, PPh2). ([R]D20 ) +202 (CHCl3, c ) 0.19)). Anal. Calcd
for C34H39N2PRu: C, 67.20; H, 6.47; N, 4.61. Found: C, 67.24;
H, 6.46; N, 4.45.
Ra cem ic (S*)-(R*)-[(Cp *F eC5H3(CH(Me)P Cy2)(P P h 2)-
1,2)P d (η3-C3H5)]+[OTf]- ((S*)-(R*)-12). To a stirred solution
OM9508997