Preparation of pseudo-C2-symmetric P,S-hybrid ferrocenyl ligand and its application to some asymmetric reactions

Article abstract of DOI:10.1016/S1381-1169(02)00634-9

A solely planar chiral pseudo-C₂-symmetric ferrocene-based P,S-chelating 2 was prepared and applied to some asymmetric reactions. Although a poor enantioselectivity is obtained in the asymmetric allylic alkylation (AAA) of rac-8 using the Pd complex bearing the P,S-hybrid ligand 2, it is noteworthy that this result is superior to that obtained using the C₂-symmetric bisphosphine counterpart of 2. The intermolecular asymmetric heck reaction (AHR) of 11 with aryl triflates catalyzed by another palladium complex containing 2 was also carried out. Although both relatively low reactivity and poor enantioselectivity were obtained, a high regioselectivy favoring the 2,5-dihydrofuran derivative (14) over the regioisomeric 2,3-dihydrofuran derivative (15) was observed in this study.

Full text of DOI:10.1016/S1381-1169(02)00634-9

Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
Preparation of pseudo-C2-symmetric P,S-hybrid ferrocenyl ligand
and its application to some asymmetric reactions
Jahyo Kang∗, Jun Hee Lee, Kwang Soo Im
Department of Chemistry, Sogang University, Seoul 121-742, South Korea
Received 9 June 2002; accepted 9 July 2002
Abstract
A solely planar chiral pseudo-C2-symmetric ferrocene-based P,S-chelating 2 was prepared and applied to some asymmetric
reactions. Although a poor enantioselectivity is obtained in the asymmetric allylic alkylation (AAA) of rac-8 using the Pd
complex bearing the P,S-hybrid ligand 2, it is noteworthy that this result is superior to that obtained using the C2-symmetric
bisphosphine counterpart of 2. The intermolecular asymmetric heck reaction (AHR) of 11 with aryl triflates catalyzed by an-
other palladium complex containing 2 was also carried out. Although both relatively low reactivity and poor enantioselectivity
were obtained, a high regioselectivy favoring the 2,5-dihydrofuran derivative (14) over the regioisomeric 2,3-dihydrofuran
derivative (15) was observed in this study.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Asymmetric catalysis; Ferrocene; P,S-ligand; Planar chirality; Pseudo-C2 symmetry
1. Introduction
rocenylbisphosphines (R-FerroPHOS derivatives (1))
with dual planar chirality (cylindrical chirality)1 [4]
as their sole element of chirality is highly effective in
the Rh-catalyzed asymmetric hydrogenation of dehy-
droamino acids [5], in the Rh-catalyzed asymmetric
hydroboration of olefins [6], and in the Pd-catalyzed
asymmetric allylic alkylation (AAA) of allyic acetates
[7] with excellent yields and excellent enantioselec-
tivities [8].
However, there would be certain situations where
the two ligating atoms of the C2-symmetric ligand
need to be different from each other while still enjoy-
ing the topological value of C2 symmetry at the same
time. This is especially true when substrates in the pro-
The incorporation of C2 symmetry in ligand design
is the most general strategy for restricting the numbers
of diastereomeric transition states in metal-catalyzed
enantioselective processes [1] since the cornerstone
preparation of DIOP by Dang and Kagan in 1971
[2]. Thereafter, such a large number of C2-symmetric
homobidentate chiral ligands (e.g. bisphosphines and
diamines) have been prepared and tested to achieve
high reactivity as well as high selectivity in catalytic
asymmetric synthesis [3].
In line with this approach, we have recently re-
ported that a series of air-stable C2-symmetric fer-
∗
1
Corresponding author. Tel.: +82-2-705-8439;
Cylindrical molecular chirality is defined here as the virtual
fax: +82-2-701-0967.
chirality derived from C2 symmetricity of two identical planar
chiralities in a given molecule.
E-mail address: kangj@ccs.sogang.ac.kr (J. Kang).
1381-1169/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(02)00634-9

56
J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
prepared from the reaction of the corresponding alco-
hol and trifluoroacetic anhydride.
2.2. General method
All experiments involving moisture- and/or
air-sensitive compounds were performed in oven-
and/or flame-dried glassware with rubber septa under
a positive pressure of nitrogen using standard Schlenk
technique. Reactions were monitored by thin-layer
chromatography (TLC) carried out on 0.25 mm E.
Merck silica gel plates (60F-254) using UV light
as a visualizing agent and p-anisaldehyde solution,
and heat as developing agent. Flash chromatography
was performed using E. Merck 230–400 mesh silica
gel according to the procedure of Still et al. [13].
Neutral TLC plates were prepared by treating the
normal TLC plates with a mixture of triethylamine
and n-hexane (1:19 (v/v)). Melting points were de-
termined using a Thomas–Hoover capillary melting
point apparatus and are uncorrected. Optical rotations
were obtained using a Rudolph Autopol III digital
polarimeter and optical rotation data are reported as
follows: [α]TD (concentration c = g/100 ml, solvent).
HPLC separations were performed with chiral station-
ary columns purchased from Daicel using a Waters
486 tunable absorbance detector operating at 254 nm
with a Waters 746 data module. GC analyses were
performed using a CP-cyclodextrin-B-2,3,6-M-19
(25 m × 0.25 mm, 0.25 ␮m film thickness) capil-
lary GC column purchased from Chrompack with
Hewlett-Packard Model HP 5890 GC fitted with
a HP3396A integrator. Infrared spectra were ob-
tained on a Mattson Galaxy 2000 spectrometer. NMR
spectra were obtained on Varian Mercury INOVA
Fig. 1. A new P,S-hybrid ligand 2 and its C2-symmetric bisphos-
phine counterpart 1.
jected reactions may well be of potentially multiden-
tate nature in transition state. Thus, a C2-symmetric
ligand backbone equipped with strong and weak donor
heteroatomic pairs (which breaks up the C2 symmetry
of the whole structure of the ligand) is needed [9].
The purpose of present study has been to compare
the effectiveness of the new pseudo-C2-symmetric
ligand 2 with its C2-symmetric bisphosphine coun-
terpart (1) in some asymmetric reactions (Fig. 1).
This differential ligation may be especially impor-
tant when additional coordination sites on a metal
for substrates and/or incoming reactants are needed.
Most notably, to our best knowledge, our new chiral
ligand 2 is the first example of pseudo-C2-symmetric
P,S-bidentate ligand that contains the planar chirality
as the sole element of chirality (a solely planar chiral
N,S-bidentate ferrocenyloxazoline has been reported
in [10]).
2. Experimental
2.1. Materials
1
Chlorodiphenylphosphine was purchased from
Aldrich, vacuum-distilled, and stored in a refrigera-
tor under N2 atmosphere. n-Butyl lithium (Aldrich)
was assayed by titration with N-benzylbenzamide as
an indicator in THF. Triethylaluminum in toluene
(25 wt.%, Aldrich) was titrated by hydrolytic meth-
ods. bis(␩3-Allyl)di-␮-chloropalladium [11] and
bis(dibenzylideneacetone)palladium [12] were pre-
pared according to the literature procedures and stored
in a refrigerator under N2 atmosphere. rac-1,3-Diphe-
nylprop-2-enyl acetate (8) was prepared from trans-
chalcone through a conventional two-step proce-
dure. Phenyl triflate and 2-(2-naphthyl) triflate were
500 (500 MHz H, 125.7 MHz 13C, and 202.4 MHz
31P NMR) spectrometer or Gemini 300 (300 MHz
1H and 75.5 MHz 13C NMR) spectrometer. 1H
NMR spectra were referenced to tetramethylsilane
(δ = 0.00 ppm) as an internal standard and are re-
ported as follows: chemical shift, multiplicity (br:
broad, s: singlet, d: doublet, t: triplet, q: quartet,
m: multiplet). 13C NMR spectra were referenced to
the residual CDCl3 (δ = 77.0 ppm) and 31P NMR
spectra were referenced to external 85% H3PO4
(δ = 0.00 ppm). Mass spectra (VG Trio 2000, low
resolution, EI) and microanalysis data (Carlo Erba
EA 1180 elemental analyzer) were provided by the

J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
57
Organic Chemistry Research Center at Sogang
University.
2.3.2. (R,R,pS,pS)-2,2’-bis(α-Acetoxyethyl)-1,1ꢀ-
dibromoferrocene (5)
mixture of dibromodiamine (4) (1.10 g,
A
2.26 mmol) and acetic anhydride (7.39 g, 72.4 mmol)
in a 25-ml Schlenk flask was thoroughly degassed
and heated at 70 ◦C for 10 h. Then, volatiles were
removed at 40 ◦C under vacuum (0.7 mmHg). The
residue was purified by column chromatography
(5% EtOAc/n-hexane) on silica gel, which was
pre-deactivated with 5% Et3N/n-hexane, to afford the
diacetate (5) (999 mg, >85%) as an orange solid. TLC
(neutral TLC, 10% EtOAc/n-hexane): Rf = 0.10;
mp: 90–92 ◦C. [α]D27 = −39.0 (c = 0.59, CHCl3). IR
(KBr): 3093, 2985, 2953, 2937, 1728, 1454, 1376,
1235, 1175, 1115, 1021, 957, 930, 841, 819, 721,
607, 490 cm−1. MS (EI, 70 eV) m/z (relative inten-
sity): 518 (M + 2, 1.10), 515 (3.33), 458 (1.39),
455 (3.28), 345 (6.10), 285 (20.8), 171 (100), 91
2.3. Preparation of P,S-hybrid ligand
2.3.1. (R,R,pS,pS)-2,2ꢀ-bis(α-N,N-Dimethylamimo-
propyl)-1,1’-dibromoferrocene (4)
To a stirred solution of diamine (3) (2.14 g,
6.01 mmol) in dry diethyl ether (20 ml, 0.3 M) was
added dropwise n-BuLi (2.30 M in hexanes, 10.4 ml,
24.0 mmol) over 15 min at room temperature. After
several minutes, the color of the mixture changed
from orange to red. After overnight, the reaction
mixture was cooled in a dry ice–acetone bath and
a solution of 1,2-dibromotetrachloroethane (8.80 g,
27.0 mmol) in dry THF (13.5 ml) was added dropwise
via cannula over 15 min. The resulting dark brown
suspension was allowed to warm to room tempera-
ture over 90 min, stirred at room temperature for 1 h,
and then quenched with saturated Na2S2O3 solution
at 0 ◦C. The mixture was diluted with diethyl ether.
The organic layer was separated and the aqueous
layer was extracted with diethyl ether. The combined
organic layers were dried over anhydrous MgSO4,
filtered, and concentrated using a rotatory evapora-
tor. The resulting residue was purified by column
chromatography (2% EtOAc/n-hexane) on silica gel,
which was pre-deactivated with 2% Et3N/n-hexane,
to afford the diastereomerically pure dibromide (4)
(2.60 g, 84%) as a brown solid. TLC (neutral TLC,
5% EtOAc/n-hexane): Rf = 0.30; mp: 46–48 ◦C.
[α]2D7 + 115 (c = 1.0, CHCl3). IR (KBr): 2971,
2930, 2820, 2777, 1450, 1375, 1042, 990, 817, 774,
511 cm−1. MS (EI, 70 eV) m/z (relative intensity): 513
(8.85), 468 (29.4), 425 (60.9), 397 (52.0), 346 (15.1),
228 (66.7), 134 (54.5), 104 (58.7), 86 (100). 1H NMR
(CDCl3, 500 MHz): δ = 4.24 ppm (dd, JHH = 2.5,
1.0 Hz, 2H, Cp), 4.13 ppm (t, JHH = 2.5 Hz, 2H,
Cp), 4.00 ppm (dd, JHH = 2.5, 1.0 Hz, 2H, Cp),
3.50 ppm (dd, JHH = 7.5, 11.2 Hz, 2H, CHCH2CH3),
2.07–1.97 ppm (m, 2H, CHCH2CH3), 2.04 ppm (s,
12H, NCH3), 1.81–1.74 ppm (m, 2H, CHCH2CH3),
1.12 ppm (t, JHH = 7.5 Hz, 6H, CHCH2CH3). 13C
NMR (CDCl3, 125.7 MHz): δ = 86.48, 81.04, 75.93,
68.69, 67.73, 60.81, 40.71, 25.91, 12.15 ppm. Anal.
Calcd. for C20H30Br2FeN2: C, 46.72; H, 5.88. Found:
C, 46.97; H, 5.45.
1
(24.7). H NMR (CDCl3, 500 MHz): δ = 5.93 ppm
(q, JHH = 6.5 Hz, 2H, CHCH3), 4.39 ppm (dd,
JHH = 1.0, 2.5 Hz, 2H, Cp), 4.32 ppm (dd, JHH
=
1.0, 2.5 Hz, 2H, Cp), 4.18 ppm (t, JHH = 2.5 Hz,
Cp), 2.01 ppm (s, 6H, C(O)CH3), 1.62 ppm (d,
JHH = 6.5 Hz, 6H, CHCH3). 13C NMR (CDCl3,
125.7 MHz): δ = 170.0, 86.75, 80.61, 75.79, 69.84,
67.01, 66.75, 20.98, 18.78 ppm. Anal. Calcd. for
C18H26Br2FeO4: C, 41.90; H, 3.91. Found: C, 41.91;
H, 3.91.
2.3.3. (pS,pS)-1,1ꢀ-Dibromo-2,2ꢀ-di(3-pentyl)
ferrocene (6)
To a stirred solution of diacetoxyferrocene (5)
(1.04 g, 1.86 mmol) in dry dichloromethane (18.6 ml,
0.1 M) was added dropwise triethylaluminum (1.90 M
solution in toluene, 4.89 ml, 9.28 mmol) at −78 ◦C.
The reaction mixture was stirred for 60 min at
−78 ◦C. The reaction mixture was warmed to room
temperature and stirred for additional 20 min at that
temperature. The mixture was transferred into the
aqueous saturated NaHCO3 solution (10 ml) at 0 ◦C
via cannula and saturated sodium potassium tar-
trate solution (10 ml) was then added. The solvent
was removed using a rotatory evaporator and the
residue was dissolved in dry diethyl ether (15 ml).
The resulting suspension was stirred vigorously for
15 min and then acidified with 1 N HCl (10 ml).
The organic layer was separated and the aqueous
layer was extracted with diethyl ether. The com-

58
J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
bined organic extracts were washed successively
with saturated NaHCO3 solution, water, and brine.
The organic layer was dried over anhydrous MgSO4,
filtered, and concentrated using a rotatory evapo-
rator. Purification by column chromatography (2%
EtOAc/n-hexane) on silica gel afforded the dibromo-
ferrocene (6) (893 mg, 99%) as an orange solid. TLC
(5% EtOAc/n-hexane): Rf = 0.78; mp: 36–37 ◦C.
[α]2D7 = +83 (c = 1.0, CHCl3). IR (KBr): 2961,
2932, 2872, 1460, 1376, 958, 809, 498 cm−1. MS
(EI, 70 eV) m/z (relative intensity): 483 (M+, 100),
404 (11.5), 374 (18.8), 324 (4.84), 239 (4.25), 180
(16.7), 153 (29.4), 134 (19.5), 104 (25.2), 78 (11.4).
1H NMR (CDCl3, 500 MHz): δ = 4.20 ppm (t,
JHH = 2.0 Hz, 2H, Cp), 4.01 ppm (t, JHH = 2.5 Hz,
2H, Cp), 3.98 ppm (m, 2H, Cp), 2.48–2.44 ppm
(m, 2H, CHCH2CH3), 1.95–1.87 ppm (m, 2H,
CHCH2CH3), 1.64–1.53 ppm (m, 4H, CHCH2CH3),
1.52–1.43 ppm (m, 2H, CHCH2CH3), 1.04 ppm (t,
JHH = 7.5 Hz, 6H, CHCH2CH3), 0.624 ppm (t,
JHH = 7.5 Hz, 6H, CHCH2CH3). 13C NMR (CDCl3,
125.7 MHz): δ = 93.84, 80.66, 74.11, 67.93, 66.58,
37.40, 25.13, 24.88, 12.35, 8.932 ppm. Anal. Calcd.
for C20H28Br2Fe: C, 49.6; H, 5.83. Found: C, 49.8;
H, 5.76.
18.6), 589 (M+, 100), 560 (21.4), 510 (29.9), 424
(0.900), 375 (15.3), 291 (30.8), 183 (58.8). H NMR
1
(CDCl3, 500 MHz): δ = 7.62–7.59 ppm (m, 2H,
C6H5), 7.38–7.36 ppm (m, 3H, C6H5), 7.22–7.13 ppm
(m, 5H, C6H5), 4.30 ppm (d, JHH = 1.5 Hz, 1H,
Cp), 4.13 ppm (t, JHH = 2.5 Hz, 1H, Cp), 3.93 ppm
(t, JHH = 2.5 Hz, 1H, Cp), 3.86–3.85 ppm (m, 1H,
Cp), 3.83–3.82 ppm (m, 1H, Cp), 3.76–3.75 ppm
(m, 1H, Cp), 2.71–2.66 ppm (m, 1H, CHCH2CH3),
2.34–2.29 ppm (m, 1H, CHCH2CH3), 1.97–1.92 ppm
(m, 1H, CHCH2CH3), 1.82–1.78 ppm (m, 1H,
CHCH2CH3), 1.59–1.50 ppm (m, 3H, CHCH2CH3),
1.46–1.38 ppm (m, 3H, CHCH2CH3), 1.01 ppm (t,
JHH = 7.5 Hz, 3H, CHCH2CH3), 0.95 ppm (t, JHH
=
7.5 Hz, 3H, CHCH2CH3), 0.55 ppm (t, JHH = 7.5 Hz,
3H, CHCH2CH3), 0.42 ppm (t, JHH = 7.5 Hz, 3H,
CHCH2CH3). 13C NMR (CDCl3, 125.7 MHz): δ
= 140.8 ppm (d, JPC = 8.5 Hz), 137.8 ppm (d, JPC
=
7.3 Hz), 135.7 ppm (d, JPC = 22 Hz), 132.2 ppm (d,
JPC = 18.4 Hz), 129.3, 128.2 ppm (d, JPC = 8.5 Hz),
127.8 ppm (d, JPC = 6.2 Hz), 101.0 ppm (d, JPC
=
25.6 Hz), 93.11, 75.30 ppm (d, JPC = 3.6 Hz), 73.68,
73.38 ppm (d, JPC = 3.6 Hz), 71.23, 67.29 ppm (d,
JPC = 31.8 Hz), 38.36, 38.28, 30.92, 25.71 ppm (d,
JPC = 3.6 Hz), 25.38, 24.94 ppm (d, JPC = 2.4 Hz),
12.42 ppm (d, JPC = 4.9 Hz), 9.00, 8.06 ppm. 31P
NMR (CDCl3, 202.4 MHz): δ = −24.81 ppm. Anal.
Calcd. for C32H38BrFeP: C, 65.44; H, 6.50. Found:
C, 65.83; H, 6.40.
2.3.4. (pS,pS)-1-Bromo-1ꢀ-diphenylphosphino-2,2ꢀ-
di(3-penty)ferrocene (7)
To a stirred solution of dibromoferrocene (6)
(690 mg, 1.42 mmol) in dry THF (5.6 ml, 0.25 M)
was added dropwise n-BuLi (2.12 M solution in
hexanes, 0.670 ml, 1.42 mmol) at −78 ◦C. After the
reaction mixture was kept at this temperature for
1 h, chlorodiphenylphosphine (345 mg, 1.56 mmol)
was then added at −78 ◦C. The reaction mixture
was warmed to room temperature over a period of
1 h, stirred at room temperature for 1 h, and then
quenched with saturated NaHCO3 solution. The mix-
ture was extracted with diethyl ether, washed with
brine, dried over MgSO4, and filtered. After removing
the solvent, the crude product was purified by column
chromatography on silica gel (2% EtOAc/n-hexane)
to afford the pure bromophosphine (7) (812 mg,
97%) as a yellow solid. TLC (5% EtOAc/n-hexane):
Rf = 0.47; mp: 52–55 ◦C. [α]28 = −179 (c = 1.0,
CHCl3). IR (KBr): 2964, 293D4, 2874, 1463, 1438,
1378, 1168, 828, 748, 698, 513, 488 cm−1. MS
(EI, 70 eV) m/z (relative intensity): 591 (M + 2,
2.3.5. (pS,pS)-1ꢀ-Methylthio-1-diphenylphosphino-
2,2ꢀ-di(3-pentyl)ferrocene (2)
To a stirred solution of bromophosphine (7)
(568 mg, 0.964 mmol) in dry THF (3.9 ml, 0.25 M)
was added dropwise n-BuLi (2.23 M solution in
hexanes, 0.454 ml, 1.01 mmol) at −78 ◦C. After the
reaction mixture was kept at this temperature for 1 h,
dimethyl disulfide (99.9 mg, 1.06 mmol) was then
added at −78 ◦C. The reaction was slowly warmed
to room temperature over 1 h, and stirred further
1 h at this temperature. The reaction was quenched
with water. The mixture was extracted with di-
ethyl ether, washed with brine, dried over MgSO4,
and filtered. After removing the solvent, the crude
product was purified by column chromatography
(1% EtOAc/n-hexane) on silica gel to afford the
P,S-hybrid ligand 2 (432 mg, 80%) as an orange oil.
TLC (5% EtOAc/n-hexane): Rf = 0.58. [α]D28
=

J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
59
−355 (c = 0.6, CHCl3). IR (KBr): 3075, 3050,
2964, 2929, 2874, 1463, 1433, 1378, 828, 743, 698,
428 cm−1. MS (EI, 70 eV) m/z (relative intensity):
556 (M+, 100), 528 (19.7), 342 (13.7), 295 (6.07),
266 (8.53), 183 (20.4), 107 (8.49), 77 (4.59). 1H
NMR (CDCl3, 500 MHz): δ = 7.64–7.61 ppm (m, 2H,
C6H5), 7.39–7.37 ppm (m, 3H, C6H5), 7.22–7.14 ppm
(m, 5H, C6H5), 4.26 ppm (dd, JHH = 1.5, 3.5 Hz,
1H, Cp), 4.13 ppm (t, JHH = 2.5 Hz, 1H, Cp),
4.07 ppm (t, JHH = 2.5 Hz, 1H, Cp), 3.95–3.93 ppm
(m, 2H, Cp), 3.55 ppm (dd, JHH = 1.0, 2.5 Hz,
1H, Cp), 2.67–2.66 ppm (m, 1H, CHCH2CH3),
2.44–2.38 ppm (m, 1H, CHCH2CH3), 1.98–1.92 ppm
(m, 1H, CHCH2CH3), 1.90–1.86 ppm (m, 1H,
CHCH2CH3), 1.87 (s, 3H, SCH3), 1.62–155 ppm (m,
3H, CHCH2CH3), 1.50–1.43 ppm (m, 1H, CHCH-
2CH3), 1.01 ppm (t, JHH = 7.5 Hz, 3H, CHCH2CH3),
0.97 ppm (t, JHH = 7.5 Hz, 3H, CHCH2CH3), 0.585
ppm (t, JHH = 7.5 Hz, 3H, CHCH2CH3), 0.451 ppm
(t, JHH = 7.5 Hz, 3H, CHCH2CH3). 13C NMR
(CDCl3, 125.7 MHz) δ = 140.9 ppm (d, JPC = 8.5
Hz),138.1 ppm (d, JPC = 8.5 Hz), 135.6 ppm (d,
JPC = 23.3 Hz), 132.2 ppm (d, JPC = 18.4 Hz), 129.1,
monitored by TLC for the disappearance of 8. When
8 disappeared completely (ca. 20 min), the solvent
was evaporated using a rotatory evaporator and the
resulting mixture was extracted with ether. The ex-
tract was washed twice with ice-cold saturated NH4Cl
solution, dried over anhydrous MgSO4, filtered, and
concentrated using a rotatory evaporator. The re-
sulting residue was purified by column chromatog-
raphy (15% EtOAc/n-hexane) on silica gel to give
(–)-(S)-dimethyl (1,3-diphenylprop-2-enyl)malonate
(9) (318 mg, >99%) as a white solid. The absolute
configuration was confirmed definitely by compar-
ison of the optical rotation with a literature value
[14]. The enantiomeric excess was determined to
be 38.5% by HPLC analysis of the product 9 with
a chiral stationary phase column (Daicel Chiral-
pak AD column, 2-propanol/n-hexane = 1:9, flow
rate: 0.8 ml/min, (R)-isomer: 14.8 min, (S)-isomer:
20.2 min).
2.5. General procedure for the asymmetric Heck
reaction of 2,3-dihydrofuran (11)
128.1 ppm (d, JPC = 7.4 Hz), 127.8 ppm (d, JPC
=
A mixture of P,S-hybrid ligand 2 (33.4 mg,
60.0 ␮mol) and Pd(dba)2 (17.2 mg, 30.0 ␮mol) in dry
solvent (0.62 ml) was stirred at 70 ◦C under N2 at-
mosphere for 1 h, and then aryl triflate (1.00 mmol),
2,3-dihydrofuran (11) (350 mg, 5.00 mmol), and
amine (3.00 mmol) were added successively. The
mixture was then heated at the appropriate temper-
ature with vigorous stirring. After cooling to room
temperature, the crude mixture was extracted with di-
ethyl ether, washed with brine, dried over anhydrous
MgSO4, and concentrated using a rotatory evaporator.
The resulting residue was purified by column chro-
matography (2% EtOAc/n-hexane) on silica gel to af-
ford (R)-2-aryl-2,5-dihydrofuran as a pale yellow oil.
The absolute configuration was confirmed definitely
by comparison of the optical rotation with a literature
value [15,16]. The enantiomeric excess was deter-
mined by HPLC analysis of the product with a chiral
stationary phase column (2-phenyl-2,5-dihydrofuran
(14): Daicel Chiralcel OD column, n-hexane, flow
rate: 1.0 ml/min, (R)-isomer: 56.0 min, (S)-isomer:
71.26 min; 2-(2-naphthyl)-2,5-dihydrofuran (16): Dai-
cel Chiralcel OD-H column, 2-propanol/n-hexane
= 1:9, flow rate: 0.5 ml/min, (R)-isomer: 19.8 min,
(S)-isomer: 21.36 min).
4.9 Hz), 127.5, 100.9 ppm (d, JPC = 25.6 Hz), 95.79,
83.17, 73.19 ppm (d, JPC = 4.8 Hz), 72.38 ppm (d,
JPC = 4.8 Hz), 72.15, 71.03, 69.67, 67.92, 38.44 ppm
(d, JPC = 8.5 Hz), 38.06, 30.90, 25.85 ppm (d,
JPC = 3.6 Hz), 25.03 ppm (d, JPC = 23.3 Hz), 19.03,
12.44 ppm (d, JPC = 2.5 Hz), 9.11, 8.18 ppm. 31P
NMR (CDCl3, 202.4 MHz): δ = −24.42 ppm. Anal.
Calcd. for C33H41FePS: C, 71.21; H, 7.43. Found: C,
71.44; H, 7.39.
2.4. General procedure for the asymmetric allylic
alkylation of 1,3-diphenylprop-2-enyl acetate (8)
A mixture of P,S-hybrid ligand 2 (12.2 mg,
22.0 ␮mol) and [Pd(␲-allyl)Cl]2 (3.66 mg, 10.0 ␮mol)
in dry dichloromethane (1.00 ml) was stirred at
room temperature under N2 for 30 min, and the
resulting red solution was added to a mixture of
1,3-diphenylprop-2-enyl acetate (8) (252 mg, 1.00
mmol) and potassium acetate (2.00 mg, 20.0 ␮mol)
in dry dichloromethane (1.00 ml) via cannula, fol-
lowed by the addition of dimethyl malonate (396 mg,
3.00 mmol) and BSA (613 mg, 3.00 mmol). The re-
action was carried out at room temperature and

60
J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
3. Results and discussion
3.2. Asymmetric allylic alkylation
3.1. Preparation of chiral ligand
Up to now, few mixed thioether-containing lig-
ands have been applied to enantioselective allylic
alkylation reactions [20]. While the coordinating
ability of thioether donors in late transition metal
complexes is well precedented (for a review see
[21]), the stereogenic center formed at sulfur upon
coordination may be flexible by its low inversion
barrier (15–20 kcal/mol) [22]. Therefore, any attempt
to incorporate a thioether donor into a chiral ligand
must consider potential erosion of enantioselectivi-
ties as a result of sulfur inversion. With this intrinsic
limitation in mind, we wished to compare the ef-
ficiency of the pseudo-C2-symmetric ligand 2 with
its C2-symmetric bisphosphine counterpart 1 in the
asymmetric catalysis. At first, we tested 2 briefly
in the Pd-catalyzed asymmetric allylic alkylation of
rac-1,3-diphenylprop-2-enyl acetate (8), since this re-
action is not only one of the best understood catalytic
asymmetric transformations but good results also can
often be obtained using nonsymmetrical heterobiden-
tate ligands (for a recent review see [23]).
As depicted in Scheme 2, using a new palladium
complex derived from 1.0 mol% of [Pd(␲-allyl)Cl]2
and 2.2 mol% of 2, we carried out the allylic alky-
lation of rac-8 with dimethyl malonate under a set
of conditions (CH2Cl2 (0.5 M), 3 eq of BSA with
5 mol% KOAc, room temperature). After 20 min, the
reaction was complete giving the (S)-9 in a quan-
titative yield with only 38.5% e.e. Although the
enantioselectivity of 38.5% is low, it is noteworthy
that this result is superior to that (5.5% e.e.) obtained
using the C2-symmetric bisphosphine counterpart
(3-Pt-FerroPHOS (1), R = 3-pentyl) [7]. In this
Our synthesis started from the nearly enantiomer-
ically pure (>99% e.e.) C2-symmetric diamine (3),
which was prepared in two steps from 1,1ꢀ-ferro-
cenedicarboxaldehyde in excellent overall yield
[5,6,8], utilizing thiazincolidine-catalyzed asymmet-
ric Et2Zn addition developed in our laboratory [17].
In the actual event, the diamine (3) was convention-
ally dilithiated with n-BuLi in diethyl ether and sub-
sequently reacted with 1,2-dibromotetrachloroethane
to furnish the C2-symmetrical dibromoferrocene (4)
in 84% yield as a single diastereomer after simple
column chromatography (Scheme 1).
After heating a mixture of 4 and excess Ac2O
at 70 ◦C for 10 h, the resulting diacetoxyferrocene
(5) was isolated in a good yield with complete
retention of configuration on the pseudo-benzylic
position [18]. Subsequently, the diacetoxyferrocene
(5) was reacted with triethylaluminum [5,6,8] at
−78 ◦C to afford the cylindrically chiral dibromofer-
rocene (6) in a quantitative yield after column chro-
matography on silica gel. A single lithium–bromine
exchange of 6 (for a similar strategy using 1,1ꢀ-bis(tri-
n-butylstannyl)ferrocene see [19]) by treating with
1.00 eq. of n-BuLi at −78 ◦C for 60 min followed
by reaction with chlorodiphenylphosphine yielded
1-bromo-1ꢀ-diphenylphosphino-2,2ꢀ-di(3-pentyl)ferr-
ocene (7) in an excellent yield. Finally, the novel
P,S-chelating ligand 2 was obtained in a good yield
as a red oil from the reaction of 7 with 1.1 eq. of
n-BuLi at −78 ◦C for 60 min followed by trapping the
derived monolithio species with dimethyl disulfide.
Scheme 1. Reagents and conditions: (a) (i) n-BuLi, ether, RT, 12 h, (ii) (BrCCl2)2, THF, −78 ◦C to RT, 84%; (b) Ac2O, 70 ◦C, 10 h, >85%;
(c) Et3Al, CH2Cl2, −78 ◦C to RT, 99%; (d) (i) n-BuLi, THF, −78 ◦C, 1 h, (ii) ClPPh2, −78 ◦C to RT, 97%; (e) (i) n-BuLi, THF, −78 ◦C,
1 h, (ii) (MeS)2, −78 ◦C to RT, 80%.

J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
61
Scheme 2. Palladium-catalyzed asymmetric allylic alkylation of rac-8 using the P,S-hybrid ligand 2.
3.3. Asymmetric Heck reaction (AHR)
Our pseudo-C2-symmetric P,S-hybrid ligand 2
was also employed in the Pd-catalyzed AHR of
2,3-dihydrofuran (11) with aryl triflates. Initially, to
evaluate the catalytic efficiency of 2 in intermolecular
AHR (for a recent review see [25]), phenyl triflate was
reacted with 5.0 eq. of 11 in the presence of 6.0 mol%
of a palladium catalyst derived from mixing 2 and
catalyst precursor (Scheme 3).
Although various reaction parameters including
base, solvent, catalyst precursor, and reaction tem-
perature, which have been known to affect strongly
on the enantioselectivity as well as the catalytic ac-
tivity, were screened extensively, all the reactions
tested gave disappointing results and some of them
are depicted in Table 1. Thus, the reaction with
Pd(dba)2 (dba: dibenzylideneacetone) as catalyst pre-
cursor (With Pd(OAc)2 as catalyst precursor and
2, the reaction did not proceed at all regardless of
conditions) did proceed at 70 ◦C in the presence of
3.0 eq. of N,N-diisopropylethylamine as base to af-
ford (R)-2-phenyl-2,5-dihydrofuran (14) in 37% yield
with 11% e.e. accompanied with a trace amount
of the isomerized product 15 (entry 1). The best
enantioselectivity (34% e.e.) was obtained when
the reaction was carried out in THF using triethy-
lamine as base albeit the yield of 14 was still low
(entry 4).
Fig. 2. A hypothetical working model predicting the major enan-
tiomer.
case, a sterically more demanding bisphosphine 1
(MME-FerroPHOS, R = (1-methoxy-1-methyl)ethyl)
was shown to be an excellent ligand [7].
The stereochemical outcome can be rationalized
with a hypothetical working model (Fig. 2). The
enantiodifferentiation step in Pd-catalyzed allylic
alkylation is the substitution of Pd(␲-allyl) com-
plexes with incoming nucleophiles, and nucleophilic
attack occurs predominantly at the allyl terminus
from trans to the better ␲-acceptor (P ꢁ S) [24].
Since the (S)-isomer was obtained as the major
enantiomer, the reaction probably proceeds through
a M-type intermediate ‘10M’ rather than a W-type
intermediate ‘10W’.
It is the most noteworthy that the major product of
every run (Table 1) was the 2,5-dihydrofuran deriva-
Scheme 3. Asymmetric Heck reaction of 11 with PhOTf using the P,S-hybrid ligand 2.

62
J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
Table 1
Asymmetric Heck reaction of 11 with PhOTf using the P,S-hybrid ligand 2a
Entry
Base
Solvent
Temperature (◦C)
Time (h)
14 (yield % (e.e. (%),
configuration)b)
15 (yield %)
1
2
3
4
5
6
i-Pr2NEt
Et3N
Proton sponge
Et3N
i-Pr2NEt
2,6-Lutidine
Benzene
Benzene
Benzene
THF
THF
THF
70
70
60
70
70
70
22
22
72
22
24
24
37 (11, R)
32 (25, R)
11 (11, R)
34 (34, R)
34 (32, R)
34 (24, R)
trace
trace
–
trace
–
–
a All reactions were carried out in 1.0 M solution using 5.0 mmol of 11, 1.0 mmol of PhOTf, and 3.0 mmol of base under N2 atmosphere,
in the presence of 3.0 mol% of Pd-catalyst derived from mixing 6.0 mol% of 6 with 3.0 mol% of Pd(dba)2, unless otherwise stated. Yields
were based on PhOTf employed.
b Enantiomeric excesses were determined by HPLC analysis using a chiral stationary column (Daicel Chiralcel OD-H column). And
absolute configurations were confirmed by comparison of the optical rotation with a literature value [15].
Scheme 4. Asymmetric Heck reaction of 11 with 2-naphthyl triflate using the P,S-hybrid ligand 2.
tive 14, although the results showed that the e.e. value
and the chemical yield were not strongly dependent
on the base and the solvent used. The correspond-
ing 2,3-dihydrofuran derivative 15, which is the major
product in the Pd(BINAP)-catalyzed reaction (BINAP:
2,2ꢀ-bis(diphenylphosphino)-1,1ꢀ-binaphthyl) [26,27],
was not detected or was formed in only a negligible
amount.
The asymmetric Heck reaction between 11 and
2-naphthyl triflate was also carried out as shown in
Scheme 4. Switching to the sterically more bulky
2-(2-naphthyl) triflate showed a positive effect on the
chemical yield of the 2-ary-2,5-dihydrofuran deriva-
tive 16 but little effect on the enantioselectivity of
the reaction. Thus, the reaction of 2,3-dihydrofuran
(11) with 2-naphthyl triflate in THF or benzene at
70 ◦C in the presence of a base such as Et3N or
2,6-lutidine gave (R)-2-(2-naphthyl)-2,5-dihydrofuran
(16) in 50–64% yield with 36–40% e.e. Somewhat
high e.e. of 40% was obtained when the reaction was
performed in benzene although an appreciable amount
of the isomerized product 17 was accompanied in this
solvent.
4. Conclusion
As described above, we have prepared a new solely
planar chiral ferrocenyl P,S-bidentate ligand 2 in good
overall yield (ca. 55% for five steps) using a pre-
dictable asymmetric methodology. The chiral ligand
2 is the first example of P,S-chelating ligand that con-
tains two different planar chiralities as the sole chiral
element. The palladium complex bearing 2 is such
a highly active catalyst for allylic alkylation that the
reaction of rac-8 with dimethyl malonate gives the
alkylation product 9 in a nearly quantitative yield only
after 20 min at room temperature. Although the enan-
tioselectivity of (S)-9 attained in this reaction is quite
low, it is somewhat superior to that obtained using
the C2-symmetric counterpart of 2. The stereochem-
ical outcome of the reaction can be easily explained
by a conventional transition state model (Fig. 2).
The intermolecular asymmetric Heck reaction of
2,3-dihydrofuran (11) with aryl triflates catalyzed by
another palladium complex containing 2 was also car-
ried out. Unfortunately, both relatively low reactivity
and poor enantioselectivity were obtained in this study.

J. Kang et al. / Journal of Molecular Catalysis A: Chemical 196 (2003) 55–63
63
However, noteworthy is that the high regioselectivy
favoring the 2,5-dihydrofuran derivative 14 over the
regioisomeric 2,3-dihydrofuran derivative 15 is ob-
served. Application of our new chiral ligand 2 to other
transition-metal-catalyzed asymmetric reactions and
further investigation with new pseudo-C2-symmetric
ligands are now in active progress in our laboratory.
[11] Y. Tatsuno, T. Yoshida, Seiotsuka, Inorg. Synth. 19 (1979)
220.
[12] T. Ukai, H. Kawazawa, Y. Ishii, J.J. Bonnett, J.A. Ibers, J.
Organomet. Chem. 65 (1974) 253.
[13] W.C. Still, M. Kahn, A. Mita, J. Org. Chem. 43 (1978) 2923.
[14] P. von Matt, A, Pfaltz, Angew. Chem., Int. Ed. Engl. 32
(1993) 566.
[15] A. Pfaltz, O. Loiseleur, M. Hayashi, N. Schmees, Synthesis
(1997) 1338.
[16] F. Ozawa, A. Kubo, T. Hayashi, Tetrahedron Lett. 33 (1992)
1485.
Acknowledgements
[17] J. Kang, J.W. Kim, J.W. Lee, D.S. Kim, J.I. Kim, Bull. Korean
Chem. Soc. 17 (1996) 1135 (and references therein).
[18] G. Gokel, P. Hoffmann, H. Klusacek, D. Marquarding, E.
Ruch, I. Ugi, Angew. Chem., Int. Ed. Engl. 9 (1970)
64.
This work was supported by grant no. 1999-1-
12300-001-5 from the Basic Research Program of the
Korea Science & Engineering Foundation.
[19] W. Zhang, Y.-I. Yoneda, T. Kida, Y. Nakatsuji, I. Ikeda,
Tetrahedron: Asym. 9 (1998) 3371.
[20] D.A. Evans, K.R. Campos, J.S. Tedrow, F.E. Michael, M.R.
Gagne, J. Am. Chem. Soc. 122 (2000) 7905 (and references
therein).
References
[21] S. Murray, F. Hartley, Chem. Rev. 81 (1981) 365.
[22] E. Abel, S.K. Bhargava, K.G. Orrell, Prog. Inorg. Chem. 32
(1984) 1.
[23] A. Pfaltz, M. Lautens, in: E.N. Jacobsen, A. Pfaltz, H.
Yamammoto (Eds.), Comprehensive Asymmetric Catalysis,
vol. II, Springer, New York, 1999, Chapter 24, p. 833.
[24] B. Akermark, B. Krakenberg, S. Hansson, A. Vitagliano,
Organometallics 6 (1987) 620.
[1] J.K. Whitesell, Chem. Rev. 89 (1989) 1581.
[2] T.P. Dang, H.B. Kagan, J. Am. Chem. Soc. 94 (1972) 6429.
[3] R. Noyori, Asymmetric Catalyst in Organic Synthesis, Wiley,
New York, 1994.
[4] E.L. Eliel, S.H. Wilen, N.N. Mander, Stereochemistry of
Organic Compounds, Wiley, New York, 1994.
[5] J. Kang, J.H. Lee, S.H. Ahn, J.S. Choi, Tetrahedron Lett. 39
(1998) 5523.
[25] M. Shibasaki, E.M. Vogl, in: E.N. Jacobsen, A. Pfaltz, H.
Yamammoto (Eds.), Comprehensive Asymmetric Catalysis,
vol. I, Springer, New York, 1999, Chapter 14, p. 457.
[26] F. Ozawa, A. Kubo, T. Hayashi, J. Am. Chem. Soc. 113
(1991) 1417.
[27] A. Miyashita, A. Yasuda, H. Takaya, K. Toriumi, T. Ito,
T. Souchi, R. Noyori, J. Am. Chem. Soc. 102 (1980)
7932.
[6] J. Kang, J.H. Lee, J.B. Kim, G.J. Kim, Chirality 12 (2000)
378.
[7] J. Kang, J.H. Lee, J.S. Choi, Tetrahedron: Asym. 12 (2001)
33.
[8] J.H. Lee, Ph.D. Thesis, Sogang University, 2000.
[9] R. Stranne, J.-L. Vasse, C. Moberg, Org. Lett. 3 (2001) 2525.
[10] S.-L. You, Y.-G. Zhou, X.-L. Hou, L.-X. Dai, Chem.
Commun. (1998) 2765.