Regioselectivity with hemispherical chelators: Increasing the catalytic efficiency of complexes of diphosphanes with large bite angles

Article abstract of DOI:10.1002/anie.200601978

(Chemical Equation Presented) Control through molecular pockets: Calixarene-derived diphosphites provide effective hemispherical crowding about catalytic centers. When used in the palladium-catalyzed alkylation of cinnamyl acetate, regioselectivities higher than 98% in favor of the linear product are observed. In the rhodium-catalyzed hydroformylation of styrene, these diphosphites lead to phenylpropionaldehyde with selectivities as high as 76%.

Full text of DOI:10.1002/anie.200601978

Communications
Diphosphite Ligands
DOI: 10.1002/anie.200601978
Regioselectivity with Hemispherical Chelators:
Increasing the Catalytic Efficiency of Complexes
of Diphosphanes with Large Bite Angles**
David SØmeril, Catherine Jeunesse, Dominique Matt,*
and Loïc Toupet
The entrapment of substrates in molecular pockets that
incorporate a metal ion constitutes an exciting approach to
catalytic transformations displaying shape-, regio-, and enan-
tioselectivity.[1–3] The environment of a bound substrate can be
closely controlled by confinement of the metal center, in
particular enabling a catalytic reaction to be driven towards
the product that best adapts to the steric restraints imposed by
the confining structure. Although these and other effects of
coordination are seen to operate ubiquitously in biology, their
transposition to catalytic systems of industrial relevance still
remains limited.[4–8]
Calix[4]arenes are molecules readily fixed in various
conformations that provide very different relative orienta-
tions of substituent groups. These substituent groups may
contain donor atoms and the commonly encountered cone
conformer thus provides not only a particularly useful plat-
[*] Dr. D. SØmeril, Dr. C. Jeunesse, Dr. D. Matt
Laboratoire de Chimie Inorganique MolØculaire
LC3 CNRS
UniversitØ Louis Pasteur
67008 Strasbourg Cedex (France)
Fax: (+33)390-241-749
E-mail: dmatt@chimie.u-strasbg.fr
Dr. L. Toupet
Groupe Mati›re CondensØe et MatØriaux
UMR 6626 CNRS
UniversitØ de Rennes 1
35042 Rennes Cedex (France)
[**] Financial support from the Agence Nationale de la Recherche is
gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
5810
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5810 –5814

Angewandte
Chemie
1
form for the construction of chelators with a constrained bite
angle,[10] but also a convenient means of generating secondary
coordination spheres.[9] We nowreport the synthesis and
characterization of pocket-shaped, calixarene-based diphos-
phites that combine the advantage of an effective bite-angle
control with that of steric pressure exerted, in the present
case, by lateral pendent groups, anchored to two distal oxygen
atoms of the macrocyclic skeleton. The catalytic activity of
PdII and RhI complexes of these diphosphites has been
evaluated in the allylic alkylation of unsymmetrical allyl
acetates and in the hydroformylation of olefins.
Diphosphites 3 and 4 were prepared by double deproto-
nation of the corresponding dialkylated precursors 1 and 2,
respectively, followed by reaction with [(S)-(1,1’-binaphtha-
lene-2,2’-diyl)]chlorophosphite (Scheme 1). The conical struc-
ture of the resulting calixarenes was unambiguously inferred
from the 13C NMR spectra, which both display ArCH2Ar
signals near 32 ppm (d = 31.96/33.30 (3), 31.78/33.54 ppm (4)),
that is, in a range typical for this conformation. In keeping
with C2 symmetry, the ArCH2 protons of each ligand appear
as two distinct AB patterns in the H NMR spectrum, while
the corresponding 31P NMR spectra exhibit a single signal in
the phosphite region (d = 133.76 (3), 123.59 ppm (4)). The
rather large separation between the distal phenoxy rings of
the calixarene moiety, combined with the steric bulk created
by the “binaphthyl” moieties, makes these ligands, a priori,
suitable for the synthesis of chelate complexes in which the P–
M–P bite angle should significantly surpass 908.
Complex 5 was obtained quantitatively by addition of one
equivalent of 3 to a [{Pd(h3-allyl)Cl}2]/NH4PF6 suspension in
acetone. The formation of a chelate complex was inferred
from the corresponding mass spectrum, which shows an
intense peak at 1507.57, corresponding to the [Pd(allyl)·3]+
cation. Owing to the presence of a h3-bonded allyl ligand, the
whole complex is no longer C2 symmetrical. This is best seen
in the 1H NMR spectrum, which exhibits four AB systems for
the ArCH2 protons. The same features were observed for the
related complex 6, obtained in a similar way from 4. As
revealed by a single crystal X-ray diffraction study, diphos-
phite 3 displays a bite angle of ca. 107.58 in complex 5.
Obviously, the steric encumbrance of the binaphthyl units
largely contributes to such a large bite angle (Figure 1), as in a
Figure 1. Molecular structure of 5. In the CPK representation (right),
the allyl group is in yellow. Selected bond lengths [Š] and angles [8]:
Pd–P1 2.292(2), Pd–P2 2.294(2); P1–Pd–P2 107.50(8), O1–P1–Pd
119.6(2), O2–P2–Pd 119.3(2); interplanar angle between the two
naphthyl moieties embracing the allyl group: 49.68; dihedral angles
between facing phenoxy rings: 81.18 (O1/O2), 2.08 (O3/O4). The PF6
anion and the six CH2Cl2 solvent molecules are omitted for clarity.
related calixarene bearing significantly smaller P(OR)2 moi-
eties the P–M–P angle was found to be only 1018.[11] The
molecular structure further shows that the metal–allyl moiety
sits deep in a pocket generated by two nearly symmetrically
sited naphthyl moieties and the two propoxy groups. We note
also that, owing to the presence of a typically flattened cone
conformation (Figure 1), the O-propyl groups are pushed
towards the palladium center, and are therefore significantly
involved in an effective embrace of the allyl group.
The ligands 3 and 4 were assessed in the Pd-catalyzed
alkylation of 3-phenylallyl acetate 7 with dimethyl malonate
(Scheme 2). With both ligands, the linear product 9 was
formed with over 98% selectivity (98.1% for 3; 98.3% for 4).
This remarkably high regioselectivity sharply contrasts with
that obtained for the same substrate using other diphos-
phanes.[12–16] Diphenylphosphanylethane (dppe), for example,
led under similar conditions to a 9:8 ratio of 91.8:8.2. The
highest linear:branched ratio obtained with a diphosphite,
Scheme 1. Synthesis of the diphosphites and their cationic allyl–
palladium complexes.
Angew. Chem. Int. Ed. 2006, 45, 5810 –5814ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5811

Communications
Scheme 3. Rhodium-catalyzed hydroformylation of 1-octene and sty-
rene. acac=acetylacetone.
Scheme 2. Palladium-catalyzed allylic substitution of the unsymmetri-
cal substrate 7. BSA=N,O-Bis(trimethylsilyl)acetamide.
us to test 3 in the hydroformylation of the less reactive trans-
2-octene. Here the TOF value droped to 170, but the
selectivity towards the linear aldehyde remained good, the
l:b ratio being kept in the range 4.3–4.6 (Table 1, entry 3). For
[24]
93:7, was reported by Claver et al. for a furanoside diphos-
phite.[16] The unusual selectivities obtained with 3 and 4 might
be explained by the presence of pockets created around the
metal center. Molecular models clearly showthat in the allyl
complex formed with 7, the unsubstituted allylic carbon atom
is the one for which the incoming nucleophile interferes least
with the naphthyl and O-alkyl groups, whatever the orienta-
tion of the phenylallyl group, syn or anti. It should be recalled
here that steric parameters become particularly important in
alkylation reactions that involve a late transition state. It is
further interesting to note that, unlike the situation found in
related Pd–allyl–Xantphos complexes,[17] the allyl embrace by
the calixarene–diphosphites involves not only P substituents,
but also auxiliary side groups axially positioned with respect
to the metal center, hence resulting in an increased steric
protection of the allyl moiety.
The homogeneous hydroformylation of olefins is the
favorite application for diphosphanes with large bite
angles.[5,18–23] The ligands 3 and 4 were therefore tested in
the rhodium-catalyzed hydroformylation of n-octenes
(Scheme 3). A large excess of phosphite was required to
obtain high regioselectivities (compare entries 1 and 2 and 4
and 5 in Table 1), indicating relatively inefficient complex-
ation and the possible involvement of a Rh catalyst species
with only one or even no phosphite coordinated through its
P atom.
comparison, using PPh3
or van Leeuwenꢀs most bulky
Xantphite[25] gave l:b ratios of only 0.9 and 2.8, respectively.
The relevance of the pendent side groups for the
regioselectivity was shown by replacing the propoxy sub-
stituents by the bulkier benzyloxy groups. Thus, in the
hydroformylation of 1-octene with 4 at 808C, the l:b ratio
rose to 80 without loss of activity (Table 1, entry 5) and with
little isomerization. When used in the hydroformylation of
trans-2-octene at 1208C, 4 resulted in a high selectivity
towards the linear aldehyde, the l:b ratio in this case reaching
24.9 (Table 1, entry 6). The latter result implies that under
these conditions, isomerization towards 1-octene occurs.
Finally, the two diphosphites were assessed in the hydro-
formylation of styrene (Table 2, entries 1 and 3). In all
catalytic runs, the linear aldehyde was the major product,
whereas conventional mono- and diphosphanes usually give
more branched product.[26] For example, at 20 bar, and after
20% conversion, the selectivity towards the linear aldehyde
was 58.8% with 3 and 73.6% with the bulkier diphosphite 4.
Decreasing the pressure of CO/H2 from 20 to 10 bar increased
the selectivity to 67.4 (3) and 76.1% (4), respectively, without
reducing significantly the activity (Table 2, entries 2 and 4).
This unusual regioselectivity can, in line with van Leeuwenꢀs
findings made for ligands with large bite angles, be assigned to
a pocket effect. The fact that the selectivity of 4 towards the
linear aldehyde is higher than that of the propoxy-substituted
ligand 3 is of course a clear indication that the degree of steric
Diphosphite 3, bearing two pendant propoxy groups,
turned out to be very effective for
the hydroformylation of 1-octene.
Table 1: Rhodium-catalyzed hydroformylation of n-octenes using diphosphites 3 and 4.
TOF values up to 1440 (Table 1,
entry 2) were observed, for a l:b
ratio as high as 58.0. We observed
that the high preference for the
linear product was maintained until
the end of the reaction. Neverthe-
less, an important proportion of
isomeric products was produced
(ca. 50% of the amount of octene
converted). This is possibly due to
the steric encumbrance created
about transient “Rh(isoalkyl)” moi-
eties, which favors b elimination in
these intermediates (and hence iso-
merization) over a carbonylation
step. The ability of the Rh/3
Entry Olefin
Ligand L/Rh t [h]
Conversion [%][a] TOF[b] l:b[c]
Isomerization [%][d]
1[e]
2[e]
1-octene
1-octene
3
3
1
10
3.75 32.1
1.2 34.5
430
1440 58.0
620 56.3
170
70
1150
1190
3.2
9.6
17.0
46.7
1.9
3.4
2.6
8
2
98.5
41.6
70.0
23.1
23.9
87.3
44.6
3[f]
trans-2-octene
3
10
4.3
4.6
7.2
8
1
4[e]
5[e]
1-octene
1-octene
4
4
1
10
1
24
8
n.d.[g] 3.0
180 80.1
50 24.9
12.1
3.3
6[f]
trans-2-octene
4
10
[a] Determined by GC; calibration based on decane. [b] mol(converted octene) per mol(Rh) and hour.
[c] Ratio, percent of linear (l) aldehyde to percent of branched (b) aldehyde. [d] Isomerized 1-octene/all
octenes or 2-octene/all octenes. [e] Reaction conditions: 1-octene (10 mmol), 1-octene/Rh=5000,
initial pressure (at 808C): 20 bar CO/H2 (1/1), T=808C, toluene/n-decane (15 mL/0.5 mL), incubation
overnightat80 8C. [f] Reaction conditions: trans-2-octene (3.2 mmol), trans-2-octene/Rh=800, initial
pressure (at120 8C): 20 bar CO/H2 (1/1), T=1208C, toluene/n-decane (15 mL/0.5 mL), incubation
system to isomerize 1-octene led overnightat80 8C. [g] Notdetermined, because of a very low amountof branched aldehyde.
5812
www.angewandte.org ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2006, 45, 5810 –5814

Angewandte
Chemie
À3
Table 2: Rhodium-catalyzed hydroformylation of styrene using diphosphites 3 and 4.[a]
0.248,
Sw = 0.938,
D1 < 2.9 eŠ .
CCDC 297616 contains the supplemen-
tary 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.
Entry
Ligand
p(CO/H2) [bar]
t [h]
Conversion [%][b] TOF[c]
Branched [%]
Linear [%]
1
2
3
3
20
10
5
5
24
4.5
5.5
24
18.3
21.7
58.7
19.1
21.4
64.4
180
220
120
210
200
134
41.2
34.8
32.6
26.4
23.9
23.4
58.8
65.2
67.4
73.6
76.1
76.6
3
4
4
4
20
10
Received: May 18, 2006
Published online: August 3, 2006
[a] Reaction conditions: styrene (10 mmol), styrene/Rh=5000, L/Rh=10, CO/H2 (1/1) initial pressure
(at80 8C), T=808C, toluene/n-decane (15 mL/0.5 mL), incubation overnight. [b] Determined by GC,
calibration based on decane. [c] mol(converted styrene) per mol(Rh) and h.
Keywords: allylic alkylation ·
.
calixarenes · diphosphites ·
hydroformylation · regioselectivity
protection about the metal center plays a key role for the
catalytic outcome. This is also in keeping with the observation
[1] B. Kersting, Z. Anorg. Allg. Chem. 2004, 630, 765.
that the selectivity obtained with 4 is somewhat better than
that reported for the Xantphites, which have similar large bite
angles, but which are deprived of any side group able to exert
an additional steric pressure on the first coordination sphere
of the metal.[25]
[2] C. Jeunesse, D. Armspach, D. Matt, Chem. Commun. 2005, 5603.
[3] T. S. Koblenz, H. L. Dekker, C. G. de Koster, P. W. N. M. van
Leeuwen, J. N. H. Reek, Chem. Commun. 2006, 1700.
[4] M. T. Reetz, S. R. Waldvogel, Angew. Chem. 1997, 109, 870;
Angew. Chem. Int. Ed. Engl. 1997, 36, 865.
[5] R. Paciello, L. Siggel, M. Röper, Angew. Chem. 1999, 111, 2045;
Angew. Chem. Int. Ed. 1999, 38, 1920.
[6] V. F. Slagt, J. N. H. Reek, P. C. J. Kamer, P. W. N. M. van Leeu-
wen, Angew. Chem. 2001, 113, 4401; Angew. Chem. Int. Ed. 2001,
40, 4271.
[7] D. Armspach, D. Matt, F. Peruch, P. Lutz, Eur. J. Inorg. Chem.
2003, 805.
[8] S. Steyer, C. Jeunesse, J. Harrowfield, D. Matt, Dalton Trans.
2005, 1301.
[9] Calixarenes 2001 (Eds.: Z. Asfari, V. Böhmer, J. M. Harrowfield,
J. Vicens), Kluwer Academic Publishers, Dordrecht, 2001.
[10] P. Kuhn, C. Jeunesse, D. SØmeril, D. Matt, P. Lutz, R. Welter, Eur.
J. Inorg. Chem. 2004, 4602.
[11] C. Kunze, D. Selent, I. Neda, M. Freytag, P. G. Jones, R.
Schmutzler, W. Baumann, A. Börner, Z. Anorg. Allg. Chem.
2002, 628, 779.
Overall, calixarenes 3 and 4 provide unprecedented
examples of hemispherical chelators able to drive allylic
alkylation as well as olefin hydroformylation reactions
towards the formation of “linear” products. With these
ligands, the metal embrace is particularly effective, the large
bite angles being associated with a conformation of the
calixarene which produces a steric protection by the O-alkyl
substituents of the apical positions on the metal, already
encumbered by the substituents on the bound P atoms. These
steric effects are cooperative consequences of the known fact
that flattening of the orientation of two opposed phenyl rings
of a calix[4]arene to the mean macrocyclic plane results in a
steeper orientation of the other pair.
[12] R. PrØtôt, A. Pfaltz, Angew. Chem. 1998, 110, 337; Angew. Chem.
Int. Ed. 1998, 37, 323.
Experimental Section
[13] S.-L. You, X.-Z. Zhu, Y.-M. Luo, X.-L. Hou, L.-X. Dai, J. Am.
Chem. Soc. 2001, 123, 7471.
[14] D. Laurenti, M. Feuerstein, G. P›pe, H. Doucet, M. Santelli, J.
Org. Chem. 2001, 66, 1633.
[15] M. DiØguez, O. Pàmies, C. Claver, J. Org. Chem. 2005, 70, 3363.
[16] M. DiØguez, O. Pàmies, C. Claver, Adv. Synth. Catal. 2005, 347,
1257.
[17] M. Kranenburg, P. C. J. Kamer, P. W. N. M. van Leeuwen, Eur. J.
Inorg. Chem. 1998, 25; Xantphos: 9,9-dimethyl-4,5-bis(diphe-
nylphosphanyl)xanthene.
[18] C. P. Casey, G. T. Whiteker, M. G. Melville, L. M. Petrovich,
J. A. Gavney, Jr., D. R. Powell, J. Am. Chem. Soc. 1992, 114,
5535.
[19] 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.
[20] P. W. N. M. van Leeuwen, P. C. J. Kamer, J. N. H. Reek, P.
Dierkes, Chem. Rev. 2000, 100, 2741.
[21] W. Ahlers, M. Röper, P. Hofmann, D. C. M. Warth, R. Paciello,
WO 01/58589 (BASF), 2001.
Full experimental details for all products are given in the Supporting
Information.
3: 1H NMR (300 MHz, C6D6): d = 7.69–7.27 (15H, CHarom), 7.18–
6.98 (10H, CHarom), 6.95–6.75 (7H, CHarom), 5.10 and 3.52 (AB sys-
tem, 4H, ArCH2Ar, 2J = 13.0 Hz), 5.00 and 3.00 (AB system, 4H,
ArCH2Ar, 2J = 13.0 Hz), 4.00–3.79 (m, 4H, OCH2), 2.06–1.85 (m, 4H,
CH2CH3), 1.26 (s, 18H, C(CH3)3), 1.23 (s, 18H, C(CH3)3), 0.39 ppm (t,
6H, CH2CH3, 3J = 7.4 Hz). 13C{1H} NMR (75 MHz, C6D6): d =
154.10–121.98 (Carom), 77.34 (s, OCH2), 33.81 (s, C(CH3)3), 33.79 (s,
C(CH3)3), 33.30 (s, ArCH2Ar), 31.96 (s, ArCH2Ar), 31.43 (s, C-
(CH3)3), 31.40 (s, C(CH3)3), 22.96 (s, CH2CH3), 9.44 ppm (s, CH2CH3).
31P{1H} NMR (121 MHz, C6D6): d = 133.76 ppm (s, OP(OAr)2).
Crystal structure of 5·6CH2Cl2: Mr = 2163.56, monoclinic, P21,
a = 13.4071(7), b = 23.4689(8), c = 17.0844(8) Š, b = 102.469(4)8, V=
5248.8(4) Š3, Z = 2, Dx = 1.369 mgmÀ3, l(MoKa) = 0.71073 Š, m =
5.91 cmÀ1, F(000) = 2228, T= 100(1) K. Data were collected on an
Oxford Diffraction Xcalibur Saphir 3 diffractometer (graphite MoKa
radiation, l = 0.71073 Š). The structure was solved with SIR-97[27]
which revealed the non-hydrogen atoms of the molecule. After
anisotropic refinement, many hydrogen atoms were found with a
Fourier difference analysis. The whole structure was refined with
SHELX-97[28] and full-matrix least-square techniques (use of F2; x, y,
z, bij for Pd, P, F, Cl, C and O atoms, x, y, z in riding mode for H atoms;
1163 variables and 11795 observations with I > 2.0 s(I); calc w = 1/
[s2(Fo2) + (0.18P)2 + 28.7P] where P = (Fo2 + 2Fc2)/3. R = 0.088, Rw =
[22] J. R. Briggs, G. T. Whiteker, Chem. Commun. 2001, 2174.
[23] R. P. J. Bronger, P. C. J. Kamer, P. W. N. M. van Leeuwen, Orga-
nometallics 2003, 22, 5358.
[24] L. A. van der Veen, P. C. J. Kamer, P. W. N. M. van Leeuwen,
Angew. Chem. 1999, 111, 349; Angew. Chem. Int. Ed. 1999, 38,
336.
Angew. Chem. Int. Ed. 2006, 45, 5810 –5814ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5813

Communications
[25] C. B. Dieleman, P. C. J. Kamer, J. N. H. Reek, P. W. N. M.
van Leeuwen, Helv. Chim. Acta 2001, 84, 3269; Xantphite
ligands are derived from the 2,7,9,9-tetramethyl-9H-xanthen-
4,5-diol skeleton and the appropriate phosphorochloridite.
[26] C. K. Brown, G. Wilkinson, J. Chem. Soc. A 1970, 2753.
[27] “SIR97: a newtool for crystal structure determination and
refinement”: A. Altomare, M. C. Burla, M. Camalli, G. Cascar-
ano, C. Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G.
Polidori, R. Spagna, J. Appl. Crystallogr. 1998, 31, 74 – 77.
[28] G. M. Sheldrick, SHELX-97, Program for the refinement of
Crystal Structures, University of Göttingen, Germany, 1997.
5814
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5810 –5814