Catalysis by design: Wide-bite-angle diphosphines by assembly of ditopic ligands for selective rhodium-catalyzed hydroformylation

Article abstract of DOI:10.1002/anie.200701255

The assembly makes the bite! Ditopic ligands comprising an anionic N-O or N-N moiety (shown in blue and gray) and a phosphine moiety (orange) can be assembled with hard metals such as zinc(II) (green) to form bidentate phosphine ligands. This approach led to a rhodium catalysts for the selective hydroformylation of 1-octene with linear-to-branched ratios up to 21:1 and rates comparable to covalently bound wide-bite-angle diphosphine ligands. (Figure Presented).

Full text of DOI:10.1002/anie.200701255

Angewandte
Chemie
DOI: 10.1002/anie.200701255
Hydroformylation Catalysts
Catalysis by Design: Wide-Bite-Angle Diphosphines by Assembly of
Ditopic Ligands for Selective Rhodium-Catalyzed Hydroformylation**
David Rivillo, Henrik Gulyµs, Jordi Benet-Buchholz, Eduardo C. Escudero-Adµn,
Zoraida Freixa, and Piet W. N. M. van Leeuwen*
Dedicated to Süd-Chemie on the occasion of its 150th anniversary
Diphosphines are powerful ligands in homogeneous cataly-
sis.[1] Until a decade ago, the ligands, with very few excep-
tions,[2] contained a covalently bonded backbone holding the
two phosphorus donors together. More recently, bidentate
diphosphines have been assembled by strong[2] or weak[3]
hydrogen-bond interactions, assembly metal interactions,[4]
and ionic interactions.[5] All of these approaches are well-
known in supramolecular chemistry and have recently been
used in organometallic complexes.[6] Also, monophosphorus
ligand systems without predetermined interactions turned out
to be extremely effective in a number of cases,[7] as did mixed
monophosphorus ligands.[8] A modular approach using mon-
odentate ligands and their connectors[4,8,9] gives access to
large libraries of new and potentially selective catalysts.
Many catalytic reactions carried out using catalysts with
bidentate phosphorus ligands are highly sensitive to the bite
angle of the diphosphine. There are rhodium-catalyzed
hydroformylation,[10] nickel- and palladium-catalyzed hydro-
cyanation,[11] and palladium-catalyzed cross-coupling reac-
tions[12] in which either the rate or the selectivity was
considerably improved by the use of, for instance, Xantphos
(4,5-bis(diphenylphosphino)-9,9-dimethyl-xanthene), which
has a natural bite angle of approximately 1108.[13] In some
cases, hydrogen-bonded bidentate ligand systems may suffice
to give results that equal or surpass those of covalently
bonded bidentate diphosphines.[2,3]
together with the hard metal (which can form a tetrahedral,
square-planar, octahedral, or (bi)pyramidal coordination
sphere) might give rise to a large variety of ligands, thus
requiring a minimum of synthetic effort. In view of the strong
bonds involved, these molecules should be more amenable to
design by molecular mechanics than those containing weaker
hydrogen bonds or nondirectional ionic interactions.
To this end, we synthesized the ditopic ligands 1–6
(Scheme 1). Ligands 1–3 are based on 3-diphenylphosphino-
2-hydroxy-5-methylbenzaldehyde 13, reported without the 5-
methyl group by Börner and co-workers,[4a,b] who used it
(assembled with an asymmetric diamine and Ti(OiPr)4 to give
Salenophos) for asymmetric rhodium-catalyzed hydroformy-
lation.
For the synthesis of 1–3, the appropriate amines were
treated with 13. The ditopic ligands 4 and 6 are made by
condensation of the corresponding aldehyde and 3-diphenyl-
phosphinoaniline, and 5 is synthesized from 3-(diphenylphos-
phino)benzaldehyde. In general, the Schiff base condensation
reactions were carried out in toluene heated at reflux in the
presence of molecular sieves. The assembled bidentate
phosphine ligands 7–12 were prepared by reaction of 1–6
with [Zn{N(SiMe3)2}2] or Ti(OiPr)4 (see the Supporting
Information). Molecular modeling calculations showed that
in particular the assemblies 10–12 might be interesting as
wide-bite-angle ligands. On the one hand, this approach
allows screening of large numbers of catalysts, but on the
other hand, sophisticated guesses must be made; indeed, 7–9
are less promising, as we will see below, both for structural
and for reactivity reasons. Note that 8 and 9 are isostructural
with SPANphos (SPANphos = 4,4,4’,4’6,6’-hexamethylspiro-
2,2’-bichroman-8,8’-bis(diphenylphosphino)), which gave sev-
eral wide-bite-angle complexes[14] in which the spiro carbon
atom fragment (CH2)2C(O)2 has been replaced by the zinc
Herein, we focus on modular ditopic ligands that contain
an anionic site and a soft donor phosphorus atom. The former
will be used to bind a hard metal for assembling the bidentate
phosphine ligand, while the latter binds a soft metal involved
in catalysis.[4] We use this approach to prepare wide-bite-angle
diphosphine ligands. The two fragments of the ditopic ligand,
[*] D. Rivillo, Dr. H. Gulyµs, Dr. J. Benet-Buchholz, E. C. Escudero-Adµn,
Dr. Z. Freixa, Prof. P. W. N. M. van Leeuwen
Institute of Chemical Research of Catalonia (ICIQ)
Av. Països Catalans 16, 43007 Tarragona (Spain)
Fax: (+34)977-920-221
=
fragment ( NR)2Zn(O)2. For several of these bidentate
phosphine assemblies, crystal structures were obtained.
Figure 1 shows the X-ray structure of 12.[15]
The coordination geometry of the zinc center in 12 is a
distorted tetrahedron. The distance between the two phos-
phorous atoms is 7.7 Š, but this can be shortened without
raising the energy by rotating the phenyl groups. The natural
bite angle calculated by molecular mechanics (MM) for a
complex with a rhodium or palladium center without config-
urational preference was approximately 1108. The corre-
sponding P···P separations found in the preliminary crystal
structures for 10 and 11 are 8 and 6.5 Š,[16] but these show the
same flexibility as 12. Preliminary crystal structures of 7 and 8
E-mail: pvanleeuwen@iciq.es
Homepage: http://www.iciq.es/english/grups_eng/vanleeuwen/
entrada.htm
[**] Spanish MEC is kindly acknowledged for a “Ramón y Cajal” contract
(Z. Freixa), project CTQ2005-03416/BQU and Consolider Ingenio
2010 (Grant N CSD2006_0003). Dr. Cristina JimØnez, Josep Maria
López, Dr. Jonathan Barr, Susana Delgado and Enrique Cequier are
acknowledged for technical support.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 7247 –7250
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7247

Communications
Figure 1. ORTEP plot of the molecular structure of 12 in two different
projections. Thermal ellipsoids are set at the 50% probability level;
hydrogen atoms are omitted for clarity. Selected interatomic separa-
tions [Š] and bond angles [8]: P1–P2 7.7; N1-Zn1-O1 97.45(8), O1-Zn-
O2 117.20(8), N1-Zn1-N2 118.82(8), N2-Zn1-O2 98.05(8).
To study the effect of the assembly of monophosphine
ligands 1–6 into diphosphine ligands 7–12 on the catalytic
properties, all ligands were used in the rhodium-catalyzed
hydroformylation of 1-octene (Table 1). Optimal phosphine-
to-rhodium ratios (at 1 mm Rh) are different for mono- and
diphosphines, and convenient average concentrations were
used.[17] An incubation time of 3 h was used before 1-octene
was added to the catalyst solution under pressure, as in situ IR
spectroscopy measurements showed that in most cases the
formation of [HRh(CO)2(phosphine)2] from [Rh(CO)2(acac)]
and the ligands tookabout 3 h at 80 8C and 20 bar of syn gas.
Ditopic ligands 1–3 and their assemblies 7–9 gave poor
Table 1: Rhodium-catalyzed hydroformylation of 1-octene.[a]
Entry Ligand
L
L/
Conv. Linear
l/b Isom. TOF
[%][c] ”10À3[d]
Rh [%][b]
aldeh. [%][c]
1
2
3
4
5
6
7
8
9
10
11
12
13
4
4
4
20
2
10
20
2
10
4
20
2
100
100
98
59
70
76
79
68
71
81
62
68
74
84
65
97
2.9 20
1.5
1.2
0.9
0.9
0.9
0.8
0.8
1.4
1.2
1.5
0.9
1.8
0.8
3.2
5.1
6.1
8
9
8
10
10
5
11
11
6
91
Scheme 1. Synthesized ligands.
100
100
92
100
100
100
95
3.4 12
3.5
9
13[e] 13
2.8 16
3.6 13
4.9 11
show P···P distances of 6 and 8 Š, but since these compounds
are more rigid, it may be more difficult to form bidentate
complexes. Notably, complex 7 shows trans nitrogen ligands,
which is not the case for the titanium complex reported by
Börner and co-workers,[4b] as the nitrogen atoms form part of
a 1,2-diphenyl-1,2-diaminoethane bridge. The new diphos-
phine ligands 7–12 react with a variety of common rhodium
and palladium precursors ([Rh(nbd)]+, [Rh(acac)],
[PdMeCl]; nbd = 2,5-nobornadiene, acac = acetylacetonate)
to both cis and trans complexes, which will be reported
elsewhere.[16]
6
12
12
PPh3
10
20
21
2.5
54
12
9
0.5
[f]
98
–
Xantphos[g] 2.2
[a] Conditions: T=808C, CO/H2 =1, P(CO/H2)=10 bar, [Rh]=1 mm in
toluene, substrate/Rh=670. [b] Percentage of 1-octene converted deter-
mined at 2 h of reaction. [c] Percentage of linear aldehyde and isomer-
ization from the total of reaction products calculated after 2 h.
[d] Average turnover frequency in molmolÀ1 hÀ1 determined at 40%
alkene conversion. [e] l/b ratio evolves during reaction (25–11; 0.4–3 h).
[f] Data from reference [3a]. [g] Data from reference [10].
7248 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 7247 –7250

Angewandte
Chemie
catalysts (not shown in Table 1), most likely because of
formation of salen complexes or phosphinophenolate com-
plexes of rhodium. In situ IR spectroscopy of 4 showed bands
at 2074 and 2009 cmÀ1, which indicate the presence of a
dicarbonyl rhodium(I) species, in accordance with literature
data of similar complexes.[18]
Keywords: ditopic ligands · homogeneous catalysis ·
hydroformylation · supramolecular catalysis · wide bite angles
.
[1] Z. Freixa, P. W. N. M. van Leeuwen, Dalton Trans. 2003, 1890 –
1901.
[2] P. W. N. M. van Leeuwen, C. F. Roobeek, R. L. Wife, J. H. G.
Frijns, J. Chem. Soc. Chem. Commun. 1986, 31 – 33.
Ligands 4–6 show catalytic performance close to that of
PPh3 (Table 1, entry 12) in terms of both selectivity and rate.
Isomerization decreases when more ligand is used (Table 1,
entries 1 and 2), and the linear/branched ratio goes up.
Consistently, ligands 4–6 give higher l/b ratios than PPh3 (e.g.
3.4 for Table 1, entry 5 vs. 2.5 for entry 12), which may point
to an intramolecular interaction between the two salen or
salen-like fragments through hydrogen-bond interactions.
The assembled bidentate diphosphines 10–12 with their
calculated natural bite angles of 110–1208 indeed gave
higher l/b ratios than the corresponding monodentate ligands.
The initial linear/branched ratio for 11 was as high as 25
(Table 1, entry 7). Ligand 12 gave a more stable system with a
l/b ratio of 21. The rates are about half those of the
monodentate ligands, which is also true for Xantphos
(Table 1, entry 13) compared to PPh3; Xantphos typically
gives higher l/b ratios (25–60). Isomerization remains relativly
high for 10–12, which may be due to the low pressure applied
or to the presence of other RhI species. Indeed, the in situ
high-pressure IR spectra of complexes of 10 show the same
species as with 4 in low concentrations. The in situ IR spectra
of 6 and 12 both show the characteristic bands for a
[HRh(CO)2(arylphosphine)2] species at 2044, 1987 (broad,
two peaks), and 1956 cmÀ1 (6 shows a few more absorptions),
but in the spectrum of 12, the intensity of the bisequatorial
diphosphine species (2044 and 1987 cmÀ1) is considerably
stronger (fulfilling our expectations), at the expense of the
equatorial–apical species (1987 and 1956 cmÀ1). These find-
ings nicely explain the increased preference for the formation
of linear aldehyde with ligand 12.[17]
In summary, we have extended the number of ditopic
ligands that bind with their hard donor atoms N and O to an
assembly metal (herein zinc and titanium) and to a soft metal
such as rhodium(I) through their P donor atom. While the
class of compounds is not new, we have shown that by proper
selection, catalytically active and selective species can be
generated in a facile manner. MM calculations have shown
that assembly by tetrahedral zinc centers may lead to wide-
bite-angle diphosphine ligands. It was found that the assem-
blies indeed give high selectivities for linear product in the
rhodium-catalyzed hydroformylation of 1-octene. The
method presented herein is extremely versatile, as both the
building blocks of the ditopic ligands and the assembly metal
can be varied extensively. The synthesis of the diphosphine
ligands usually involves only three steps. Furthermore, the
assembly metal fragments can be modified by additional
donor molecules or additional anionic fragments, which could
be used to make the ligands chiral.
[3] a) B. Breit, W. Seiche, J. Am. Chem. Soc. 2003, 125, 6608 – 6609;
b) L. K. Knight, Z. Freixa, P. W. N. M. van Leeuwen, J. N. H.
Reek, Organometallics 2006, 25, 954 – 960; c) P. A. Duckmanton,
A. J. Blake, J. B. Love, Inorg. Chem. 2005, 44, 7708 – 7710.
[4] a) A. Kless, R. Kadyrov, A. Boerner, J. Holz, H. B. Kagan,
Tetrahedron Lett. 1995, 36, 4601 – 4602; b) A. Kless, C. Lefeber,
A. Spannenberg, R. Kempe, W. Baumann, J. Holz, A. Boerner,
Tetrahedron 1996, 52, 14599 – 14606; c) J. M. Takacs, D. S.
Reddy, S. A. Moteki, D. Palencia, H. Wu, J. Am. Chem. Soc.
2004, 126, 4494; d) D. H. Leung, R. G. Bergman, K. N. Ray-
mond, J. Am. Chem. Soc. 2007, 129, 2746 – 2747; e) X. Sun, D. W.
Johnson, D. L. Caulder, R. E. Powers, K. N. Raymond, E. H.
Wong, Angew. Chem. 1999, 111, 1386 – 1390; Angew. Chem. Int.
Ed. 1999, 38, 1303 – 1307; f) F. Lam, J. X. Xu, K. S. Chan, J. Org.
Chem. 1996, 61, 8414 – 8418; g) S. Chikkali, D. Gudat, M.
Niemeyer, Chem. Commun. 2007, 981 – 983.
[5] H. Gulyµs, J. Benet-Buchholz, E. C. Escudero-Adµn, Z. Freixa,
P. W. N. M. van Leeuwen, Chem. Eur. J. 2007, 13, 3424 – 3430.
[6] T. S. Koblenz, H. L. Dekker, C. G. de Koster, P. W. N. M. van
Leeuwen, J. N. H. Reek, Chem. Commun. 2006, 1700 – 1702.
[7] a) C. Claver, E. Fernandez, A. Gillon, K. Heslop, D. J. Hyett, A.
Martorell, A. G. Orpen, P. G. Pringle, Chem. Commun. 2000,
961 – 962; b) M. van den Berg, A. J. Minnaard, E. P. Schudde, J.
van Esch, A. H. M. de Vries, J. G. de Vries, B. L. Feringa, J. Am.
Chem. Soc. 2000, 122, 11539 – 11540; c) M. T. Reetz, G. Mehler,
Angew. Chem. 2000, 112, 4047 – 4049; Angew. Chem. Int. Ed.
2000, 39, 3889 – 3890.
[8] a) M. T. Reetz, T. Sell, A. Meiswinkel, G. Mehler, Angew. Chem.
2003, 115, 814 – 817; Angew. Chem. Int. Ed. 2003, 42, 790 – 793;
b) D. Peæa, A. J. Minnaard, J. A. F. Boogers, A. H. M. de Vries,
J. G. de Vries, B. L. Feringa, Org. Biomol. Chem. 2003, 1, 1087 –
1089; c) M. T. Reetz, X. Li, Angew. Chem. 2005, 117, 3019 – 3021;
Angew. Chem. Int. Ed. 2005, 44, 2959 – 2962; d) M. Kuil, P. E.
Goudriaan, P. W. N. M. van Leeuwen, J. N. H. Reek, Chem.
Commun. 2006, 4679 – 4681; e) V. F. Slagt, P. W. N. M. van Leeu-
wen, J. N. H. Reek, Angew. Chem. 2003, 115, 5777 – 5781; Angew.
Chem. Int. Ed. 2003, 42, 5619 – 5623.
[9] X.-B. Jiang, L. Lefort, P. E. Goudriaan, A. H. M. de Vries,
P. W. N. M. van Leeuwen, J. G. de Vries, J. N. H. Reek, Angew.
Chem. 2006, 118, 1245 – 1249; Angew. Chem. Int. Ed. 2006, 45,
1223 – 1227.
[10] M. Kranenburg, Y. E. M. van der Burgt, P. C. J. Kamer,
P. W. N. M. van Leeuwen, Organometallics 1995, 14, 3081 – 3089.
[11] M. Kranenburg, P. C. J. Kamer, P. W. N. M. van Leeuwen, D.
Vogt, W. Keim, J. Chem. Soc. Chem. Commun. 1995, 2177 – 2778.
[12] a) M. C. Harris, O. Geis, S. L. Buchwald, J. Org. Chem. 1999, 64,
6019 – 6022; b) Y. Guari, D. S. van Es, J. N. H. Reek, P. C. J.
Kamer, P. W. N. M. van Leeuwen, Tetrahedron Lett. 1999, 40,
3789 – 3790.
[13] C. P. Casey, G. T. Whiteker, Isr. J. Chem. 1990, 30, 299 – 304.
[14] C. JimØnez-Rodríguez, F. X. Roca, C. Bo, J. Benet-Buchholz,
E. C. Escudero-Adµn, Z. Freixa, P. W. N. M. van Leeuwen,
Dalton Trans. 2006, 268 – 278.
[15] X-ray structure determination: Crystals of 12 were obtained by
slow diffusion of cyclohexane into dichloromethane at room
temperature. Although the measured crystal was stable under
atmospheric conditions, it was prepared under inert conditions
and immersed in perfluoropolyether for manipulation. Measure-
ments were made on a Bruker–Nonius diffractometer equipped
Received: March 21, 2007
Published online: August 6, 2007
Angew. Chem. Int. Ed. 2007, 46, 7247 –7250
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7249

Communications
with an APPEX 2 4 K CCD area detector, a FR591 rotating
peak(hole) = 0.568 (À0.835) eŠÀ3. CCDC-641173 contains the
supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[16] D. Rivillo, H. Gulyµs, J. Benet-Buchholz, E. C. Escudero-Adµn,
Z. Freixa, P. W. N. M. van Leeuwen, unpublished results.
[17] a) P. W. N. M. van Leeuwen, C. P. Casey, G. T. Whiteker,
Catalysis by Metal Complexes, Vol. 22, Kluwer, Dordrecht,
2000, chap. 4, pp. 63 – 105; b) L. A. van der Veen, P. H. Keeven,
G. C. Schoemaker, J. N. H. Reek, P. C. J. Kamer, P. W. N. M.
van Leeuwen, M. Lutz, A. L. Spek, Organometallics 2000, 19,
872 – 883.
anode with MoKa radiation, Montel mirrors as monochromator,
and a Kryoflex low-temperature device (T= À1738C). Full-
sphere data collection was used with w and f scans. Programs
used: Data collection Apex2V. 1.0–22 (Bruker-Nonius 2004),
data reduction Saint + Version 6.22 (Bruker-Nonius 2001), and
absorption correction SADABS V. 2.10 (2003). For structure
solution and refinement, SHELXTL Version 6.10 (Sheldrick,
2000) was used. Crystal data for 12 at 100 K: C50H38N2O2P2Zn1,
826.13 gmolÀ1
15.8553(14),
,
orthorhombic, space group Pbca, a =
b = 18.3592(18), c = 26.862(2) Š, V=
7819.2(12) Š3, Z = 8, 1calcd = 1.404 MgmÀ3, R1 = 0.0495 (0.1014),
wR2 = 0.1059 (0.1278), for 7205 reflections with I > 2s(I) (for
11564 reflections [Rint: 0.1609] with a total measured of 91491
reflections), goodness-of-fit on F2 = 1.000, largest difference
[18] J. H. H. Ho, D. St Clair Black, B. A. Messerle, J. K. Clegg, P.
Turner, Organometallics 2006, 25, 5800 – 5810.
7250
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 7247 –7250