SPANphos: A C2-symmetric trans-coordinating diphosphane ligand

Article abstract of DOI:10.1002/anie.200390330

Exclusively trans coordination is exhibited by SPANphos, a C₂-symmetric diphosphane ligand containing a spiro center, which can be easily synthesized in three steps from cheap reagents. The crystal structures of the ligand and of the complex [P

Full text of DOI:10.1002/anie.200390330

Communications
an overview on the vast number of known diphosphanes
revealed that only a limited number of trans-chelating
diphosphanes has been reported, and even fewer asymmetric
variants thereof.[1] Diphosphanes that coordinate only in a
trans fashion that is not achieved through coordination of an
additional heteroatom (i.e., in fact tridentate ligands), are
very rare. The TRANSphos ligand, developed by Venanzi
et al.,[2] is a flexible ligand that can achieve chelation angles of
175.78, but cis structures have also been reported. In 1991, Ito
et al. introduced a new family of trans-diphosphanes, namely,
the TRAP ligands,[3] which already proved to be active in
several asymmetric catalytic reactions in which cis-coordinat-
ing ligands failed to give enantioselectivity.
The synthesis of new, purely trans-diphosphanes is still a
worthwhile challenge, because they will provide access to
scarcely explored catalysts with unusual dispositions of
ligands around the active metal center. By using trans-
chelating ligands, some coordination positions that are tradi-
tionally available when cis-diphosphanes are used can be
inhibited, and new families of catalysts with new properties
could thus be discovered.[1]
We report here the synthesis of a new trans-coordinating
diphosphane, which is readily available in three steps from
cheap reagents, characterization of Rh, Pd, and Pt complexes
thereof, and some preliminary catalytic experiments.
trans-Coordinating Diphosphane
SPANphos: A C2-Symmetric trans-Coordinating
Scheme 1 shows the synthetic path to SPANphos (3).
Racemic 4,4,4’,4’,6,6’-hexamethylspiro-2,2’-bichroman (1)
was obtained on a 10 g scale by acid-catalyzed reaction of p-
cresol with acetone.[4] Spirane 1 was brominated with N-
bromosuccinimide (NBS) to obtain 8,8’-dibromo-4,4,4’,4’,6,6’-
hexamethylspiro-2,2’-bichroman (2), which was treated with
nBuLi and chlorodiphenylphosphane to give SPANphos in
about 80% overall yield. Optically pure samples of 3 were
obtained by racemic resolution by chiral HPLC.[5]
Diphosphane Ligand**
Zoraida Freixa, Mark S. Beentjes, Guido D. Batema,
Cedric B. Dieleman, Gino P. F. van Strijdonck,
Joost N. H. Reek, Paul C. J. Kamer, Jan Fraanje,
Kees Goubitz, and Piet W. N. M. van Leeuwen*
In the last few decades much effort has been devoted to the
development of sophisticated ligands for specific applications.
Among these, chelating diphosphanes could be considered to
be the most popular class of ligands for catalysis. Surprisingly,
The molecular structure of 3 was determined by X-ray
crystallographic analysis (Figure 1).[6] It has an intramolecular
P···P distance of 4.991 Š. All atoms in each chroman fragment
are nearly in a single plane, except for the spiro carbon atom
Br
PPh2
1) nBuLi
acetone, H+
OH
NBS
2) PClPh2
O
O
O
O
O
O
Br
PPh2
1
2
SPANphos(3)
Scheme 1. Synthetic route to SPANphos 3.
on the idealized C2 axis, which lies about 0.7 Š above either
plane. The dihedral angle between the two mean planes of
43.9(1)8 results in an “anti” disposition of the lone pairs of the
P atoms, which seems incompatible with a chelating coordi-
nation mode of the ligand.
[*] Prof. Dr. P. W. N. M. van Leeuwen, Dr. Z. Freixa, M. S. Beentjes,
G. D. Batema, Dr. C. B. Dieleman, Dr. G. P. F. van Strijdonck,
Dr. J. N. H. Reek, Dr. P. C. J. Kamer, J. Fraanje, K. Goubitz
Institute of Molecular Chemistry, University of Amsterdam
Nieuwe Achtergracht 166, 1018 WV Amsterdam (The Netherlands)
Fax: ( +31)20-525-6456
Reaction of
3
with stoichiometric amounts of
E-mail: pwnm@science.uva.nl
[PtCl2(CH3CN)2] and [PdCl2(cod)] (cod = 1,5-cycloocta-
diene) gave trans-[PtCl2(3)] (4) and trans-[PdCl2(3)] (5).
[**] The NRSCC is kindly acknowledged for financial support.
1284
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4211-1284 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 11

Angewandte
Chemie
Figure 1. Molecular structure of 3 (ORTEP plot).
Figure 2. Molecular structure of 4 (ORTEP plot). Selected bond
Figure 2 shows the molecular structure of 4.[7] The structure
confirms the trans coordination of the diphosphane ligand,
with a P-Pt-P angle of 171.98. This angle is about 108 wider
than those reported for square-planar Pd and Rh complexes
of TRAP derivatives.[8] This wider bite angle is likely related
to the larger P···P distance enforced by the bichroman
backbone, as opposed to the highly congested biferrocenyl
backbone of TRAP ligands. In contrast to the free ligand, in
complex 4, the oxygen atom and the carbon atom adjacent to
the spiro center lie above and below the plane defined by the
aromatic ring of the chroman fragment. This brings the
phosphorous atoms 0.40 Š closer to one another than in the
free ligand (P···P in 4: 4.591 versus 4.991 Š in the free ligand)
and appropriately orientates the lone pairs of these atoms for
trans chelation.
À
À
lengths [Š] and angles [8] for 4. Pt1 Cl1 2.294(8), Pt1 Cl2 2.338(9),
À
À
Pt1 P1 2.315(5), Pt1 P2 2.293(4), Cl1-Pt1-Cl2 176.4(3); Cl1-Pt1-P1
89.3(2), Cl1-Pt1-P2 92.9(2), Cl2-Pt1-P1 91.1(3), Cl2-Pt1-P2 86.2(2),
P1-Pt1-P2 171.9(4).
nals are consistent both with a mononuclear complex
[RhH(CO)2(3)] (9) and a binuclear species 10 with two
bridging diphosphane ligands. Both spectra are consistent
with bisequatorial coordination of the P-donor ligand, as was
corroborated by HP-IR spectroscopy under CO/H2 and CO/
D2. Further investigations to distinguish between these two
species by using DOSY techniques are planned.
The catalytic activity of metal complexes containing rac-3
was tested in several reactions. It was shown to be an efficient
ligand for rhodium-catalyzed hydroformylation of 1-octene
and styrene, palladium-catalyzed allylic alkylation of 1,3-
The 31P NMR spectrum of 4 shows a single signal with Pt
1
satellites (d = 17.6 ppm, J(Pt,P) = 2723 Hz), which strongly
suggests that the mononuclear trans complex is the only
species in solution. The 31P NMR spectrum of 5 also shows
a single signal with the same chemical shift as the Pt
complex, which indicates the same molecular structure for
both complexes. Furthermore, the related [PdClMe(3)]
(6) shows a singlet at d = 24.7 ppm in the 31P NMR
H3C
P
O
O
P
[D8] toluene
[Rh(acac)(CO)2]
3
+
Rh
HC
2
spectrum and a triplet (d = À0.27 ppm, J(P,H) = 6.1 Hz)
CO
1
in the H NMR spectrum, indicative of the trans disposi-
7
H3C
tion of the bidentate ligand.[9]
The reaction of [Rh(acac)(CO)2] with a stoichiometric
quantity of 3 produced the mononuclear complex [Rh-
(acac)(CO)(h1-3)] (7; Scheme 2); this indicates that the
ligand cannot achieve a 908 chelation angle. This was also
confirmed when complex 7 was treated with 20 bar of CO
in a high-pressure (HP) NMR tube. Formation of the new
binuclear complex [Rh2(m-CO)2(CO)4(h1-3)2] (8) was
observed, as demonstrated by the 31P NMR signals (d =
21.1, m, 1P; À15.6 ppm, brs, 1P) as well as by HP-IR
experiments. Complex 8 slowly evolves under 20 bar of H2
into a new species characterized by a doublet (d =
33.3 ppm, 1J(Rh,P) = 143.7 Hz) in the 31P NMR spectrum
H
20 bar CO
P
Rh
OC
P
CO
O
C
OC
P
P
9
or
20 bar H2
OC
Rh
Rh
CO
CO
H
H
P
P
C
O
P
P
P
P
Rh
Rh
OC
CO
8
CO
CO
10
1
and a triplet of doublets in the H NMR spectrum (d =
Scheme 2. Rhodium complexes of 3, which were confirmed by NMR and IR
1
1
À9.5 ppm, J(Rh,P) = 6.2, J(Rh,H) = 3.7 Hz). These sig-
experiments.
Angew. Chem. Int. Ed. 2003, 42, No. 11
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4211-1285 $ 20.00+.50/0
1285

Communications
CH34b’); 13C NMR (125.7 MHz, CDCl3): d = 144.94, 133.53, 131.86,
(CH arom.), 131.09, 126.15, (CH arom.), 111.22 (CH arom.), 98.60
(C2), 46.86 (CH23, CH23’), 33.07 (CH3), 32.21 (CH3), 31.18 (C4, C4’),
20.54 ppm (CH3).
diphenylprop-2-ene-1-yl acetate and Michael reaction of
ethyl a-cyanopropionate with methyl vinyl ketone.
No activity was observed for palladium-catalyzed CO/
ethene oligomerization under different conditions, and not
even the formation of methyl propanoate was detected.
Recent results suggest that a cis arrangement of the diphos-
phane ligand is necessary to enable solvolysis of the acyl
group, which is the main chain-transfer mechanism in CO/
ethene oligomerization cycles starting with a palladium
hydride species.[10] Hence, these results reinforce the idea
that 3 neither forms cis compounds, nor undergoes dissoci-
ation of one of the phosphorus atoms during the catalytic
cycle.
In conclusion, we have synthesized a new C2-symmetric
trans-diphosphane ligand that is readily accessible from cheap
reagents. Neither cis–trans isomerization nor decoordination
of one of the P atoms appears to occur under CO/ethene
oligomerization conditions. In Pd or Pt complexes SPANphos
acts as a purely trans-coordinating diphosphane ligand. This is
corroborated by NMR studies on rhodium complexes, which
showed that chelation angles close to 908 cannot be achieved
by this ligand.
As we mentioned before, the introduction by Ito et al. of
the TRAP type ligands showed that these chiral trans-
coordinating diphosphanes provide excellent enantiomeric
excesses in some reactions in which cis-diphosphanes are
inefficient.[3] From the viewpoint of enantioselective discrim-
ination, the main difference between cis-and trans-diphos-
phanes is the role of their chiral backbone. For instance, in a
square-planar complex, when a trans-chelating diphosphane
is used, both the backbone and the substituents on the P
atoms directly contribute to the formation of a chiral cavity
around the metal center. This represents an advantage over
the classical cis-chelating ligands, in which the backbone is not
directly exposed to the active sites of the metal, and the chiral
environment is mainly produced by the substituents on the
phosphorus atoms by chirality-transfer mechanisms. The
potential of the chiral cavity generated by TRAP ligands is
3: nBuLi (2.5 mL, 2.5m in hexane, 1.2 ” 10À2 mol) was added
dropwise to a refluxing solution of 2 (2.99 g, 5.8 ” 10À3 mol) in THF
(150 mL). Upon completion of the reaction, monitored by GC-MS,
ClPPh2 (2.3 mL, 1.2 ” 10À2 mol) was added, and the reaction mixture
stirred at room temperature for 16 h. The reaction was quenched with
degassed water and extracted with Et2O. The combined organic layers
were dried over magnesium sulfate, and the solvent was evaporated.
1
The product was crystallized from methanol (yield 89%). H NMR
(300 MHz, CDCl3): d = 7.42–7.00 (m, 22H; CH arom), 6.64 (m, 2H;
CH arom.), 2.00 (s, 6H; CH36, CH36’), 1.82 (d, 2J(H,H) = 14.1 Hz,
2H; H3a, H3a’), 1.82 (s, 6H; CH34a, CH34a’),1.62 (d, 2J(H,H) =
14.1 Hz, 2H; H3b, H3b’), 1.22 ppm (s, 6H; CH34b, CH34b’):
13C NMR (75.4 MHz, CDCl3): d = 151.80 (d, J(P,C) = 10.8 Hz; C
arom.), 139.26 (d, J(P,C) = 8.4 Hz; C arom.), 138.56 (d, J(P,C) =
8.1 Hz; C arom.), 134.92 (d, J(P,C) = 12.4 Hz; CH arom.), 134.44 (d,
J(P,C) = 11.6 Hz; CH arom.), 133.24 (s; CH arom.), 131.79 (s; C
arom.), 131.25 (s; C arom.), 129.14–128.20 (m, C; CH arom.), 125.55
(d, J(P,C) = 8.8 Hz; C arom.), 98.80 (s; C2), 47.65 (s; CH23, CH23’),
34.29 (CH3), 33.08 (CH3), 30.94 (C4, C4’), 21.35 ppm (CH3). 31P NMR
(121.5 MHz; CDCl3): d = À14.6 ppm (s).
CCDC-196129 and CCDC-196130 contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the
Cambridge Crystallographic Data Centre, 12, Union Road, Cam-
bridge CB21EZ, UK; fax: (+ 44)1223-336-033; or deposit@ccdc.cam.
ac.uk).
Received: October 25, 2002 [Z50435]
Keywords: homogeneous catalysis · ligand design · platinum ·
.
P ligands
[1] a) C. A. Bessel, P. Aggarwal, A. C. Maerschilok, K. J. Takeuchi,
Chem. Rev. 2001, 101, 1031 – 1066; b) W. Winter, Angew. Chem.
1976, 88, 260; Angew. Chem. Int. Ed. Engl. 1976, 15, 241; c) A.
Kless, C. Lefeber, A. Spannenberg, R. Kempe, W. Baumann, J.
Holz, A. Bꢀrner, Tetrahedron 1996, 52, 14599 – 14606.
[2] a) N. J. DeStefano, D. K. Johnson, R. M. Lane, L. M. Venanzi,
Helv. Chim. Acta 1976, 59, 2674 – 2682; b) P. N. Kapoor, L. M.
Venanzi, Helv. Chim. Acta 1977, 60, 2824 – 2829.
À
evidenced by the remote stereoselective formation of C C
bonds on atoms not directly attached to the metal center.[11]
Investigations on SPANphos derivatives containing the
same or related backbones will lead us to a family of
potentially chiral, easily accessible trans ligands. Catalytic
experiments on asymmetric reactions with enantiopure sam-
ples of the 3 are currently under development.
[3] a) M. Sawamura, H. Hamashima, Y. Ito, Tetrahedron: Asymme-
try 1991, 2, 593; b) M. Sawamura, H. Hamashima, Y. Ito, J. Am.
Chem. Soc. 1992, 114, 8295 – 8296; c) M. Sawamura, H. Hama-
shima, Y. Ito, Tetrahedron 1994, 50, 4439 – 4454; d) R. Kuwano,
K. Sato, T. Kurokawa, D. Karube, Y. Ito, J. Am. Chem. Soc. 2000,
122, 7614 – 7615; e) R. Kuwano, M. Sawamura, Y. Ito, Bull.
Chem. Soc. Jpn. 2000, 73, 2571 – 2578; f) R. Kuwano, M.
Sawamura, J. Shirai, M. Takahashi, Y. Ito, Bull. Chem. Soc.
Jpn. 2000, 73, 485 – 496.
Experimental Section
[4] This product was obtained by an improved procedure with
respect to that already described: a) A. Ryabov, Neftekhimiya
1987, 27, 215 – 218; b) A. J. Caruso, J. L. Lee, J. Org. Chem. 1997,
62, 1058 – 1063.
Bichroman 1 was prepared according to a modified literature
procedure.[4]
2: A solution of NBS (3.23 g) in DMF (40 mL) was added
dropwise to a solution of 1 (2.56 g, 7.25 ” 10À3 mol) in DMF (125 mL).
The reaction mixture was monitored by GC-MS. After five days, it
was quenched with water and extracted with Et2O. The combined
organic layers were dried over magnesium sulfate, and the solvent was
evaporated. The product was crystallized from methanol to give a
[5] Authors are indebted to DSM Research for the enantiomeric
separation of a sample of 3.
ꢀ
[6] C47H46O2P2·CH2Cl2, Mr = 704.8, triclinic, P1, a = 11.9743(5), b =
13.240(1), c = 14.808(1 Š, a = 73.807(7), b = 79.854(8), g =
71369(7)8,
V= 2126.6(3) Š3,
Z = 2,
1calcd = 1.23 gcmÀ3
,
1
white solid (yield 84%). H NMR (500 MHz, CDCl3): d = 7.11, 7.09
l(CuKa) = 1.5418 Š, m(CuKa) = 2.37 mmÀ1, F(000) = 832, room
temperature, final R = 0.061 for 6974 observed reflections.
[7] C47H46Cl2O2P2Pt, Mr = 487.2, orthorhombic, Pbn21, a =
13.167(1), b = 17.369(1), c = 17.735(2) Š, V= 4056.0(6) Š3, Z =
(s, s, 2H, 2H; H5, H5’, H7, H7’), 2.29 (s, 6H; CH36, CH36’), 2.21 (d,
2J(H,H) = 14.0 Hz, 2H; H3a, H3a’), 2.10 (d, J(H,H) = 14.0 Hz, 2H;
H3b, H3b’), 1.78 (s, 6H; CH34a, CH34a’), 1.38 ppm (s, 6H; CH34b,
2
1286
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4211-1286 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 11

Angewandte
Chemie
4, 1calcd = 1.59 gcmÀ3, l(CuKa) = 1.5418 Š, m(CuKa) = 8.73 mmÀ1
,
F(000) = 1944, room temperature, final R = 0.059 for 3641
observed reflections.
[8] M. Sawamura, H. Hamashima, M. Sugawara, R. Kuwano, Y. Ito,
Organometallics 1995, 14, 4549 – 4558.
[9] M. A. Zuideveld, B. H. G. Swennenhuis, D. K. Maarten, Y.
Guari, G. P. F. van Strijdonck, J. N. H. Reek, P. C. J. Kamer, K.
Goubitz, J. Fraanje, M. Lutz, A. L. Spek, P. W. N. M. van Leeu-
wen, J. Chem. Soc. Dalton Trans. 2002, 11, 2308 – 2317.
[10] P. W. N. M van Leeuwen, M. A. Zuideveld, B. H. G. Swennen-
huis, Z. Freixa, P. C. J. Kamer, K. Goubitz, J. Fraanje, M. Lutz, A.
Spek, unpublished results.
[11] a) M. Sawamura, H. Hamashima, H. Shinoto, Y. Ito, Tetrahedron
Lett. 1995, 36, 6479 – 6482; b) M. Sawamura, M. Sudoh, Y. Ito, J.
Am. Chem. Soc. 1996, 118, 3309 – 3310; c) R. Kuwano, H.
Miyazaki, Y. Ito, Chem. Commun. 1998, 71 – 72.
Angew. Chem. Int. Ed. 2003, 42, No. 11
ꢀ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4211-1287 $ 20.00+.50/0
1287