Phosphino-functionalised acetals of polyvinyl alcohol as the matrix for the immobilisation of Rh-based pre-catalysts for interfacial catalysis

Article abstract of DOI:10.1016/j.molcata.2005.12.044

Use of polyvinyl alcohol (PVA) to support the immobilisation of phosphines shall be presented. Phosphino functionalisation is introduced in the polymer by a one-step transacetalation reaction and high acetalation degrees are reached. By complexation of the phosphino groups to a [(COD)RhCl] complex fragment, [Rh]-modified polymers are obtained with Rh contents of higher than 15% and the transition metal in a defined molecular environment. The phosphino or [Rh]-modified PVAs are characterised unambiguously by comparing their spectroscopic data with those of model compounds, the syntheses of which are reported as well. Use of the [Rh]-modified polymers as pre-catalysts will be presented with the hydroformylation of 1-octene being taken as an example. It leads to a selectivity in the product distribution that is identical to that of the model complexes. Recycling experiments were performed and the catalyst rest state on the polymer was identified after 10 consecutive runs. Adsorption of the [Rh]-modified PVA on an inorganic support leads to the formation of stable polymer layers on its surface, which cannot be mobilised by extraction with organic solvents. The stability of these layers, together with their tuneable polarity, allow for their application in interfacial catalysis comparable to SAPC or SILC.

Full text of DOI:10.1016/j.molcata.2005.12.044

Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
Phosphino-functionalised acetals of polyvinyl alcohol as the matrix for the
immobilisation of Rh-based pre-catalysts for interfacial catalysisଝ
Martin Ahlmann, Olaf Walter∗, Melany Frank, Wilhelm Habicht
ITC-CPV, Forschungszentrum Karlsruhe, P.O. Box 3640, 76021 Karlsruhe, Germany
Received 18 November 2004; received in revised form 18 November 2005; accepted 27 December 2005
Available online 8 February 2006
Abstract
Use of polyvinyl alcohol (PVA) to support the immobilisation of phosphines shall be presented. Phosphino functionalisation is introduced in
the polymer by a one-step transacetalation reaction and high acetalation degrees are reached. By complexation of the phosphino groups to a
[(COD)RhCl] complex fragment, [Rh]-modified polymers are obtained with Rh contents of higher than 15% and the transition metal in a defined
molecular environment. The phosphino or [Rh]-modified PVAs are characterised unambiguously by comparing their spectroscopic data with those
of model compounds, the syntheses of which are reported as well. Use of the [Rh]-modified polymers as pre-catalysts will be presented with the
hydroformylation of 1-octene being taken as an example. It leads to a selectivity in the product distribution that is identical to that of the model
complexes. Recycling experiments were performed and the catalyst rest state on the polymer was identified after 10 consecutive runs. Adsorption
of the [Rh]-modified PVA on an inorganic support leads to the formation of stable polymer layers on its surface, which cannot be mobilised by
extraction with organic solvents. The stability of these layers, together with their tuneable polarity, allow for their application in interfacial catalysis
comparable to SAPC or SILC.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Polyvinyl alcohol; Acetalation; Phosphine; Immobilisation; Interfacial catalysis
1. Introduction
This problem could be solved by the application of phases,
where the catalyst is immobilised in a stationary phase with a
tuneable polarity. It was therefore proposed to use polyvinyl
alcohol (PVA) as the matrix for the immobilisation of catalysts
(Fig. 1). PVA results from the polymerisation of vinyl acetate
with subsequent hydrolysis of the ester groups yielding a poly-
mer with 98% of head-to-tail-linked vinyl alcohol units.
PVA itself is a polymer, it is the free hydroxyl groups of which
contain a lot of reactive centres which might be funtionalised
with donor groups. The most obvious and common approach
to functionalising PVA is the transformation of two hydroxyl
groups localised in the 3-position into cyclic acetals (Fig. 1)
[6]. In their 2-position, these acetals might carry a functional-
isation in the form of a spacer to which a donor group can be
connected that is eventually binding a catalytically active tran-
sition metal fragment. Via the degree of transformation of the
hydroxyl groups into 1,3-dioxanes (ratio n/m, Fig. 1), which can
be controlled by the stoichiometry of the reaction, the polar-
ity of the functionalised PVA can be tuned. In this way, even
unpolar substrates, such as longer-chain olefins, may enter the
polymer material and be catalytically transformed. Adsorption
Modern development of catalysis involves the aspects of
green chemistry [1] and sustainability [2] or responsible care
[3]. This includes the development of new processes that were
originally based on homogeneous systems as well as catalysts
recycling. The best-known examples of catalyst recycling based
on a simple phase separation technology with the utilisation of
a polar phase and the dissolution of the catalyst are the SHOP
process [4] and the Ruhrchemie–Rhone–Poulenc process for the
hydroformylation of propene [5]. For the hydroformylation of
longer-chain olefins the Ruhrchemie–Rhone–Poulenc process is
ˆ
ˆ
not applicable because of the high polarity of the aqueous cata-
lyst phase, which leads to a reduced olefin solubility and, hence,
to a small substrate/catalyst contact. This eventually decreases
the reaction rates.
ଝ
In memory of Prof. Dr. Birgit Drießen-Ho¨lscher.
Corresponding author. Tel.: +49 7247 826509; fax: +49 7247 822244.
∗
E-mail address: olaf.walter@itc-cpv.fzk.de (O. Walter).
1381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2005.12.044

M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
81
[14], 2-(ortho-diphenylphosphinophenyl)-1,3-dioxane (1g)
[15], 2-(ortho-diisopropylphosphino-phenyl)-1,3-dioxane (1h)
[15] and {chloro(COD)[2-(ortho-diphenylphosphinophenyl)-
1,3-dioxane]rhodium(I)} (2g) [15], {chloro(COD)[2-(ortho-
diisopropylphosphinophenyl)-1,3-dioxane]rhodium(I)} (2h)
[15]. Solvents were dried and purified by standard methods.
NMR spectra were recorded using a 250 MHz Bruker AVANCE
or a Varian Unity INOVA 400 MHz spectrometer at 25 ◦C
and referenced to the residual proton signal of the solvent
CDCl3 (δ = 7.27 ppm, 1H NMR spectra) or its carbon frequency
Fig. 1. Schematic view of a representative part of polyvinyl alcohol (left) and of
polyvinyl alcohol functionalised by transacetalation (spacer, D: donor function,
MLn: transition metal fragment).
(δ = 77.5 ppm, 13C{ H} NMR spectra) or to H3PO4 as external
1
standard (31P{ H} NMR spectra). All non-proton NMR spectra
1
were registered under proton decoupling. The mass spectra
were obtained using a Hewlett-Packard LC MSD Serie 1100,
ionisation method API-ES from methanol solutions of the
compounds containing NH4OAc. Infrared spectra and FIR
spectra were recorded by a Perkin-Elmer 2000 spectrometer.
The elemental analyses of solid products were performed using
a Vario EL CHN analyser.
For the crystallographic details of the single-crystal X-ray
diffraction measurements, see Table 3. The X-ray analyses were
performed using a Siemens SMART CCD 1000 diffractome-
ter with an irradiation time of 10–20 s/frame, thus collecting a
full sphere of data using an ω-scan technique with ꢀω rang-
ing from 0.3 to 0.45◦. For searches relating to single-crystal
X-ray diffraction data, the Cambridge Structural Database was
used [16]. Crystallographic data of the structures have been
deposited at the Cambridge Crystallographic Database Centre,
supplementary publications nos. CCDC 256144 (1b), 256145
(2b), 256146 (2e), 256147 (2f), and 286209 (3f). Copies of this
information may be obtained free of charge from: The Direc-
tor, CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (fax:
+44 1223 336033, e-mail: deposit@ccdc.cam.ac.uk or www:
http://www.ccdc.cam.ac.uk).
Fig. 2. Schematic view of a PVA layer on the surface of an inorganic support
with its possible application as a support in a catalytic reaction, the catalyst being
immobilised in the polymer matrix.
of the modified PVA on an inorganic support would lead to an
additional heterogenisation of the system with layers of the mod-
ified PVA being formed on the surface of the inorganic support.
Pre-catalysts prepared in this way could be applied in interfacial
catalysis comparable to SAPC [7], SILP [8] or other supported
systems (Fig. 2) [9].
Here, it shall be reported about the introduction of phos-
phino functionalisations in PVA via transacetalation reactions
with acetals under the formation of 1,3-dioxanes on the polymer
backbone. Thecoordinationofthepolymer-boundphosphinesto
a [(COD)RhCl] complex fragment shall be presented. The poly-
mers will be characterised by comparing their spectroscopic and
catalytic behaviour with that of suitable model complexes.
2.1. Description of the general procedure for the syntheses
of the ligands 1
A bromobenzaldehyde acetal solution in THF reacted at
−78 ◦C with an equimolar amount of BuLi. After stirring
for 1 h at −78 ◦C, an equimolar amount of the corresponding
chlorodiphenyl- or chlorodiisopropylphosphine was added to
this solution and the reaction mixture was allowed to warm up
to room temperature. After hydrolysis, phase separation, and
re-extraction of the aqueous phase with 20 ml dichloromethane,
the organic solvents were removed in vacuum. Solid phosphines
were purified by recrystallisation, liquids were distilled in vac-
uum.
2. Experimental
All reactions were performed in an inert gas atmosphere
using standard Schlenk techniques. Chlorodiisopropylphos-
phine purchased from Aldrich and chlorodiphenylphosphine
from Fluka were degassed and stored under an inert atmosphere
before use. For the syntheses reported, a PVA with a molecular
˚
weight of 22,000 was used. The molecular sieve (5 A, particle
2.1.1. Amounts, yields, and spectroscopic data of the
ligands 1
size 2.4–4.8 mm for reasons of better handling) used for coating
was purchased from Carl Roth Chemicals. The following chem-
icals have already been described and were prepared according
to literature: [(COD)RhCl]2 [10], [Rh(CO)2Cl]2 [11], 2-(para-
bromophenyl)-1,3-dioxane [12], diphenylphosphine [13],
ortho-(dimethoxymethylphenyl)diphenylphosphine (1c) [14],
(p-Dimethoxymethylphenyl)(diphenyl)phosphine, 1a:
12.27 g (53 mmol) 4-bromobenzaldehydedimethylacetal in
200 ml THF. Yield: 8.250 g (25 mmol, 46%) 1a in the form of a
colourless liquid with a bp of 190 ◦C at 1 × 10−5 mbar.
IR (KBr, cm−1): 3070 (m), 3052 (m), 2951 (s), 2932 (s), 2828
(s), 1585 (m), 1479 (s), 1464 (m), 1434 (ss), 1395 (m), 1350 (s),
ortho-(dimethoxymethylphenyl)diisopropylphosphine
(1d)

82
M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
MS (m/z, %): 348 (C22H21O2P+, 100%), 290 (C19H15OP+,
16%), 262 (C18H15P+, 25%), 183 (C12H8P+, 48%), 108
(C6H5P+, 14%), 87 (C4H7O2+, 6%).
C22H21O2P (%)—calc.: C 75.85, H 6.08, found: C 75.83, H
6.23.
1103 (ss), 1089 (ss), 1055 (ss), 1027 (m), 998 (m), 811 (s), 744
(ss), 696 (ss), 505 (s).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 3.35 (s, 6H,
CH3), 5.39 (s, 1H, CHO2), 7.32–7.45 (m, 14H, Ar).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = −5.3 (s).
13C NMR (63 MHz, CDCl3, 300 K): δ (ppm) = 52.9
(s, 2C, OCH3), 103.1 (s, 1C, CHO2), 126.8–138.6
(m, 18C, Ar).
2-(para-Diisopropylphosphinophenyl)-1,3-dioxane,
1f:
4.52 g (18.6 mmol) 4-(1,3-dioxane-2-yl)bromobenzene in 50 ml
THF. Yield: 2.18 g (7.8 mmol, 42%) 1f as colourless liquid with
bp of 135 ◦C at 1 mbar.
MS (m/z, %): 336 (C21H21O2P+, 97%), 305 (C20H18OP+,
76%), 290 (C19H15OP+, 31%), 183 (C12H8P+, 100%), 108
(C6H5P+, 25%).
IR (CsI, cm−1): 3078 (w), 2962 (ss), 2951 (ss), 2925 (s), 2892
(m), 2865 (ss), 1605 (w), 1563 (w), 1467 (ss), 1428 (m), 1378
(ss), 1362 (m), 1237 (s), 1150 (ss), 1110 (ss), 1089 (m), 1019
(ss), 814 (ss), 661 (m), 610 (m).
(p-Dimethoxymethylphenyl)(diisopropyl)phosphine,
1b:
12.27 g (53 mmol) 4-bromobenzaldehydedimethylacetal in
200 ml THF. Yield: 11.72 g (44 mmol, 83%) 1b in the form of
a colourless liquid with a bp of 130 ◦C at 1 mbar.
FIR (CsI, cm−1): 483 (m), 433 (w), 385 (m), 353 (w).
1H NMR (400 MHz, CDCl3, 300 K): δ (ppm) = 0.87 (dd,
3J (H,H) = 6.9 Hz, 3J (H,P) = 11.3 Hz, 6H, CH3), 1.04
(dd, 3J (H,H) = 7.0 Hz, 3J (H,P) = 15.2 Hz, 6H, CH3),
1.43 (dpsept, 2J (H,H) = 13.4 Hz, 3J (Heq,Hax) = 2.6 Hz,
3J (Heq,Heq) = 1.3 Hz, 1H, (OCH2)2CHHeq), 2.08 (dsept,
3J (H,H) = 7.0 Hz, 2J (H,P) = 1.7 Hz, 2H, PCH), 2.22
(pqt, 2J (H,H) = 13.4 Hz, 3J (Hax,Hax) = 12.4 Hz, 3J
(Hax,Heq) = 5.0 Hz, 1H, (OCH2)2CHHax), 3.97 (ptdd, 2J
(H,H) = 10.5 Hz, 3J (Hax,Hax) = 12.4 Hz, 3J (Hax,Heq) = 2.6 Hz,
4J (Hax,Heq) = 1.7 Hz, 2H, OCHHax), 4.26 (ddpt, 2J
IR (KBr, cm−1): 3076 (w), 2986 (m), 2950 (s), 2928 (s), 2866
(s), 2828 (s), 1563 (w), 1464 (s), 1445 (s), 1396 (s), 1382 (s),
1362 (s), 1350 (s), 1103 (ss), 1057 (s), 811 (s), 661 (s), 609 (s),
547 (m), 534 (m), 501 (m).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 0.71 (dd,
3
3
3J (H,H) = 6.9 Hz, J (H,P) = 11.3 Hz, 6H, CH3), 0.88 (dd, J
(H,H) = 6.9 Hz, 3J (H,P) = 15.2 Hz, 6H, CH3), 1.91 (sept, 3J
(H,H) = 6.9 Hz, 2H, PCH), 3.14 (s, 6H, OCH3), 5.19 (s, 1H,
CHO2), 7.21–7.31 (m, 4H, Ar).
31P NMR (162 MHz, CDCl3, 300 K): δ (ppm) = 11.2 (s).
3
3
(H,H) = 10.5 Hz, J (Heq,Hax) = 5.0 Hz, J (Heq,Heq) = 1.3 Hz,
4J (Heq,Hax) = 1.7 Hz, 2H, OCHHeq), 5.49 (s, 1H, CHO2),
7.45–7.46 (m, 4H, Ar).
13C NMR (100 MHz, CDCl3, 300 K): δ (ppm) = 18.6 (d, J
2
(C,P) = 8.6 Hz, 2C, CH3), 19.8 (d, 2J (C,P) = 18.4 Hz, 2C, CH3),
22.7 (d, 1J (C,P) = 11.5 Hz, 2C, PCH), 52.8 (s, 2C, OCH3), 103.1
(s, 1C, CHO2), 126.0–138.6 (m, 6C, Ar).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 12.1 (s).
13C NMR (63 MHz, CDCl3, 300 K): δ (ppm) = 19.0 (d,
MS (m/z, %,): 268 (C15H25O2P+, 53%), 237 (C14H22OP+,
33%), 195 (C11H16OP+, 100%), 183 (C9H12O2P+, 38%), 152
(C8H9OP+, 62%).
2J (C,P) = 8.0 Hz, 2C, CH3), 20.1 (d, J (C,P) = 18.4 Hz, 2C,
2
CH3), 23.1 (d, 1J (C,P) = 11.5 Hz, 2C, PCH), 26.1 (s, 1C,
(OCH2)2CH2), 67.8 (s, 2C, OCH2), 101.8 (s, 1C, CHO2),
125.8–139.6 (m, 6C, Ar).
2-(para-Diphenylphosphinophenyl)-1,3-dioxane, 1e: 6.79 g
(27.9 mmol) 4-(1,3-dioxane-2-yl)bromobenzene in 50 ml THF.
Yield: 5.74 g (16.5 mmol, 59%) of colourless 1e, with mp of
81 ◦C.
MS (m/z, %): 280 (C16H25O2P+, 55%), 238 (C13H19O2P+,
22%), 195 (C10H12O2P+, 47%), 163 (C10H11O2+, 15%), 137
(C7H6OP+, 25%), 109 (C6H6P+, 70%), 87 (C4H7O2+, 53%),
43 (C3H7+, 100%).
IR (KBr, cm−1): 3068 (m), 3046 (m), 2972 (s), 2951 (m),
2922 (m), 2856 (s), 1605 (m), 1584 (m), 1572 (w), 1477 (s),
1466 (m), 1437 (s), 1434 (s), 1376 (s), 1148 (s), 1101 (s), 1087
(s), 1025 (m), 1020 (m), 821 (s), 745 (s), 696 (s), 553 (s) 508
(s), 491 (s).
2.2. Description of the general procedure for the syntheses
of the complexes 2
FIR (PE, cm−1): 490 (s), 470 (m), 449 (s), 415 (m), 399
(w), 394 (w), 364 (m), 377 (m), 320 (m), 281 (m), 270 (m),
256 (m).
Half of the stoichiometric amount of [Rh(COD)Cl]2 was
added to a solution of the ligands 1 in an appropriate
amount of dichloromethane. After stirring for 30 min to 1 h
at RT, the solvent was removed in vacuum. The crude
products were purified by extraction with diethylether or
pentane.
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.38
(dpsept, 2J (H,H) = 13.0 Hz, 3J (Heq,Hax) = 2.0 Hz, 3J
(Heq,Heq) = 1.0 Hz, 1H, (OCH2)2CHHeq), 2.15 (pqt, 2J
(H,H) = 13.0 Hz, 3J (Hax,Hax) = 12.5 Hz, 3J (Hax,Heq) = 5.0 Hz,
1H, (OCH2)2CHHax), 3.92 (ptd, 2J (H,H) = 12.0 Hz, 3J
(Hax,Hax) = 12.5 Hz, 3J (Hax,Heq) = 2.0 Hz, 2H, OCHHax),
4.20 (ddd, 2J (H,H) = 12.0 Hz, 3J (Heq,Hax) = 5.0 Hz, 3J
(Heq,Heq) = 1.0 Hz, 2H, OCHHeq), 5.43 (s, 1H, CHO2),
7.19–7.40 (m, 14H, Ar).
2.2.1. Amounts, yields, and spectroscopic data of the
complexes 2
{Chloro(COD)[(diisopropyl)(para-dimethoxymethylphenyl)
phosphin]rhodium(I)}, 2b: 1.46 g (5.3 mmol) 1b in 5 ml
dichloromethane. Extraction with pentane for 24 h yields 2.50 g
(91%) 2b in the form of yellow crystals.
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = −4.4 (s).
13C NMR (100 MHz, CDCl3, 300 K): δ (ppm) = 26.1 (s,
1C, (OCH2)2CH2), 67.8 (s, 2C, OCH2), 101.8 (s, 1C, CHO2),
126.4–139.6 (m, 18C, Ar).
IR (KBr, cm−1): 3052 (w), 2981 (m), 2952 (m), 2870 (m),
2831 (m), 1600 (w), 1494 (w), 1473 (m), 1462 (m), 1425 (m),

M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
83
1380 (m), 1101 (ss), 1052 (ss), 818 (m), 651 (m), 623 (m), 545
{Chloro(COD)[2-(para-diisopropylphosphinophenyl)-
(m), 520 (m).
1,3-dioxane]rhodium(I)}, 2f: 447 mg (1.6 mmol) 1f in 5 ml
dichloromethane. Extraction with pentane for 24 h yields
643 mg (1.2 mmol, 76%) 2f in the form of yellow crystals.
IR (KBr, cm−1): 3056 (w), 2970 (m), 2960 (m), 2952 (m),
2930 (m), 2913 (m), 2867 (m), 2836 (m), 1523 (w), 1475 (m),
1463 (m), 1434 (m), 1424 (m), 1382 (m), 1377 (m), 1366 (m),
1235 (s), 1152 (s), 1104 (ss), 1017 (ss), 821 (m), 653 (m), 629
(m), 537 (s).
FIR (PE, cm−1): 281 (s), ν (Rh-Cl).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.18 (dd,
3J (H,H) = 6.9 Hz, J (H,P) = 13.6 Hz, 6H, CH3), 1.46 (dd, J
(H,H) = 7.0 Hz, 3J (H,P) = 15.5 Hz, 6H, CH3), 1.86 (m, 2H, CH2
(COD)), 1.97 (m, 2H, CH2 (COD)), 2.34 (m, 4H, CH2 (COD),
2.70 (poct, 2J (H,P) = 3J (H,H) = 7.5 Hz, 2H, PCH), 3.22 (s
(br), 2H, CH (COD)trans to Cl), 3.35 (s, 6H, OCH3), 5.36 (s
(br), 2H, CH (COD)trans to P), 5.39 (s, 1H, CHO2), 7.47–7.49
(m, 4H, Ar).
3
3
FIR (PE, cm−1): 283 (ss), ν (Rh-Cl).
1H NMR (400 MHz, CDCl3, 300 K): δ (ppm) = 1.14 (dd,
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 39.4 (d, J
1
3J (H,P) = 13.6 Hz, J (H,H) = 7.0 Hz, 6H, CH3), 1.44 (dd, J
(H,P) = 15.7 Hz, 3J (H,H) = 7.2 Hz, 7H, CH3, (OCH2)2CHHeq),
1.80 (m, 2H, CH2 (COD)), 1.93 (m, 2H, CH2 (COD)),
2.24 (pqt, 2J (H,H) = 13.6 Hz, 3J (Hax,Hax) = 12.4 Hz, 3J
(Hax,Heq) = 4.9 Hz, 1H, (OCH2)2CHHax), 2.28 (m, 4H, CH2
(COD)), 2.68 (septd, 3J (H,H) = 7.1 Hz, 2J (H,P) = 1.1 Hz,
3
3
(P,Rh) = 147.0 Hz).
13C NMR (63 MHz, CDCl3, 300 K): δ (ppm) = 18.9 (s,
2
1
2C, CH3), 20.0 (d, J (C,P) = 4.6 Hz, 2C, CH3), 23.6 (d, J
(C,P) = 22.4 Hz, 2C, PCH), 28.6(s, 2C, CH2 (COD)), 32.9(s, 2C,
CH2 (COD)), 53.0 (s, 2C, OCH3), 70.2 (d, 1J (C,Rh) = 13.8 Hz,
1
2C, CH (COD)trans to Cl), 102.7 (s, 1C, CHO2), 103.1 (dd, J
(C,Rh) = 7.5 Hz, J (C,P) = 11.5 Hz, 2C, CH (COD)trans to P),
2H, PCH), 3.16 (s (br), 2H, CH (COD)trans
Cl),
to
2
3.98 (ptdd, 2J (H,H) = 10.8 Hz, 3J (Hax,Hax) = 12.4 Hz, 3J
(Hax,Heq) = 2.5 Hz, 4J (Hax,Heq) = 1.1 Hz, 2H, OCHHax),
4.26 (ddpt, 2J (H,H) = 10.8 Hz, 3J (Heq,Hax) = 4.9 Hz, 3J
126.2–139.7 (m, 6C, Ar).
ESI-MS (m/z, %): 479 (C23H37O2PRh+, 100%), 269
(C15H26O2P+, 9%).
4
(Heq,Heq) = 1.1 Hz, J (Heq,Hax) = 1.2 Hz, 2H, OCHHeq), 5.31
C23H37ClO2PRh (%)—calc.: C 53.65, H 7.24, found: C
53.70, H 7.24.
(s (br), 2H, CH (COD)trans to P), 5.49 (s, 1H, CHO2), 7.40–7.52
(m, 4H, Ar).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 40.4 (d, J
1
{Chloro(COD)[2-(para-diphenylphosphinophenyl)-1,3-
dioxane]rhodium(I)}, 2e: 361 mg (1.0 mmol) 1e in 5 ml
dichloromethane. Extraction for 2 days with diethylether yields
535 mg (0.9 mmol, 86%) 2e as yellow crystals.
IR (KBr, cm−1): 3050 (w), 2958 (m), 2936 (m), 2917 (m),
2868 (m), 2829 (m), 1481 (m), 1468 (w), 1435 (m), 1377 (m),
1149 (m), 1103 (ss), 1092 (ss), 1018 (m), 814 (m), 748 (m), 698
(ss), 527 (s).
(P,Rh) = 146.6 Hz).
13C NMR (63 MHz, CDCl3, 300 K): δ (ppm) = 19.2 (s,
2C, CH3), 20.4 (d, 2J (C,P) = 4.0 Hz, 2C, CH3), 24.1 (d,
1J (C,P) = 22.3 Hz, 2C, PCH), 26.1 (s, 1C, (OCH2)2CH2),
28.9 (d, 2J (C,Rh) = 1.7 Hz, 2C, CH2 (COD)), 33.1 (d, 2J
(C,Rh) = 2.3 Hz, 2C, CH2 (COD)), 67.8 (s, 2C, OCH2), 70.7
(d, 1J (C,Rh) = 13.8 Hz, 2C, CH (COD)trans to Cl), 101.3 (s, 1C,
FIR (PE, cm−1): 284 (s), ν (Rh-Cl).
1
2
CHO2), 103.3 (dd, J (C,Rh) = 7.5 Hz, J (C,P) = 12.1 Hz, 2C,
CH (COD)trans to P), 125.9–140.4 (m, 6C, Ar).
ESI-MS (m/z, %): 491 (C24H37O2PRh+, 100%), 281
(C16H26O2P+, 16%).
1H NMR (400 MHz, CDCl3, 300 K): δ (ppm) = 1.43
(dpsept,
2J
(H,H) = 13.6 Hz,
3J
(Heq,Hax) = 2.5 Hz,
3J (Heq,Heq) = 1.3 Hz, 1H, (OCH2)2HHax), 1.88 (m,
2H, CH2 (COD)), 2.04 (m, 2H, CH2 (COD)), 2.19
(pqt, 2J (H,H) = 13.6 Hz, 3J (Hax,Hax) = 12.4 Hz, 3J
(Hax,Heq) = 4.9 Hz, 1H, (OCH2)2CHHax), 2.37 (m, 4H,
C24H37ClO2PRh (%)—calc.: C 54.71, H 7.08, found: C
54.35, H 7.32.
CH2 (COD)), 3.11 (s (br), 2H, CH (COD)trans
Cl),
2.3. Description of the general procedure for the syntheses
of the modified PVA derivatives P1
to
3.96 (ptdd, 2J (H,H) = 10.8 Hz, 3J (Hax,Hax) = 12.4 Hz, 3J
4
(Hax,Heq) = 2.5 Hz, J (Hax,Heq) = 1.1 Hz, 2H, OCHHax), 4.24
2
3
(dd, J (H,H) = 12.5 Hz, J (Heq,Hax) = 5.0 Hz, 2H, OCHHeq),
5.49 (s, 1H, CHO2), 5.55 (s (br), 2H, CH (COD)trans to P),
7.35–7.78 (m, 14H, Ar).
An overstoichiometric amount of the phosphino-func-
tionalised benzaldehydedimethylacetals (1a–1d), together
with a stoichiometric amount of PVA, were dissolved in an
appropriate amount of NMP. Gaseous HCl was added and
the reaction mixture was warmed up to about 115 ◦C for a
minimum of 12 h. The biggest part of the solvent was removed
in vacuum, and the polymers (P1e–P1h) were precipitated
either by the addition of pentane after isolation by the addition
of basic water or by the direct addition of basic water-containing
KOH. For removal of salts, the polymers were purified by
extraction with water for several days. Subsequent extraction
with diethylether over CaH2 for several days eliminates
the remaining traces of water and yields the polymers as
colourless or slightly coloured powders after treatment in
vacuum.
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 31.5 (d, J
1
(P,Rh) = 149.8 Hz).
13C NMR (63 MHz, CDCl3, 300 K): δ (ppm) = 26.15 (s, 1C,
(OCH2)2CH2), 29.3 (s, 2C, CH2 (COD)), 33.5 (s, 2C, CH2
(COD)), 67.8 (s, 2C, OCH2), 71.1 (d, 1J (C,Rh) = 13.8 Hz,
1
2C, CH (COD)trans to Cl), 101.3 (s, 1C, CHO2), 105.5 (d, J
(C,Rh) = 5.2 Hz, 2C, CH (COD)trans to P), 126.0–140.9 (m, 18C,
Ar).
ESI-MS (m/z, %): 559 (C30H33O2PRh+, 100%), 349
(C22H22O2P+, 11%).
C30H33ClO2PRh (%)—calc.: C 60.57, H 5.59, found: C
59.41, H 5.89.

84
M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
IR (KBr, cm−1): 3060 (w), 2949 (w), 2917 (w), 2864 (w),
1116 (ss), 758 (m), 638 (m), 618 (m).
2.3.1. Amounts, yields, and spectroscopic data of the
phosphino-functionalised polymers P1e–P1h
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = −7.2 (s).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 0.9 (s (br),
CH3), 1.1 (s (br), CH3), 1.6 (s (br), CH2), 2.1 (s (br, sh), PCH),
4.1 (s (sh), OCHR, ROH), 6.3 (s (br, sh), CHO2), 6.6 (s (br, sh),
CHO2), 7.4–7.9 (m, Ar-H).
Synthesis of P1e: 528 mg (1.6 mmol) 1a and 109 mg PVA
(22.000) in 40 ml NMP. Extraction with water for 5 days,
followed by extraction with a 1:1 mixture of pentane and
diethylether for 3 days, and with diethylether over CaH2 for
3 days. After drying for 12 h in vacuum, 329 mg (88%) slightly
beige coloured P1e were isolated.
Analysis (%)—found: C 66.62%, H 6.49%.
IR (KBr, cm−1): 3069 (w), 3051 (w), 2943 (m), 2911 (m),
2860 (m), 1584 (w), 1479 (m), 1433 (s), 1116 (s), 1063 (s), 1018
(ss), 812 (s), 742 (ss), 694 (ss), 496 (s).
Acetalation degree: 85% (1H NMR), 63% (analysis).
Molar mass of monomer unit: 308 g/mol (1H NMR).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = −5.2 (s).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.7 (s (br, sh),
CHn), 2.1 (s (br), CHHax), 4.2 (s (br, sh), CHOR2, ROH), 5.5 (s
(br), CHO2), 5.8 (s (br), CHO2), 7.3 (s (br, sh), Ar-H).
Analysis (%)—found: C 74.32, H 6.05.
2.4. Description of the general procedure for the syntheses
of the Rh-modified PVA derivatives P2e–P2h
A slightly overstoichiometric amount of [Rh(COD)Cl]2 was
added to a THF solution of the polymers P1e–P1h. After stirring
for at least 12 h at room temperature, the solvent was reduced to
a volume of 5 ml and the polymers P2e–P2h were precipitated
by the addition of pentane. After washing with pentane, the Rh-
modified polymers were extracted with refluxing pentane for
several days. Drying in vacuum releases P2e–P2h.
Acetalation degree: 79% (1H NMR), 77% (analysis).
Molar mass of the monomer unit: 384 g/mol.
Synthesis of P1f: 1.34 g (5.0 mmol) 1b and 437 mg PVA in
20 ml NMP. P1f was extracted with water for 2 days and with
diethylether (over CaH2 as drying agent) for 1 day. After drying
for 24 h at 1 mbar and 80 ◦C, 865 mg (2.7 mmol, 71%) beige
coloured P1f were obtained.
IR (KBr, cm−1): 3078 (w), 3027 (w), 2949 (s), 2922 (s), 2866
(s), 1462 (m), 1380 (m), 1117 (ss), 1058 (s), 1018 (ss), 813 (s),
660 (w), 609 (w).
2.4.1. Amounts, yields, and spectroscopic data of the
Rh-modified polymers P2e–P2h
Synthesis of P2e: 182 mg (0.47 mmol) P1e and 168 mg
(0.3 mmol) [Rh(COD)Cl]2 in 50 ml THF. Yield: 233 mg
(0.37 mmol, 78%) yellow coloured P2e.
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 11.3 (s).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 0.9 (s (br),
CH3), 1.1 (s (br), CH3), 1.6 (s (br), CHn), 2.1 (s (br, sh), PCH),
4.2 (s (br, sh), OCHR, ROH), 5.6 (s (br), CHO2), 5.8 (s (br),
CHO2), 7.5 (s (br), Ar-H).
IR (KBr, cm−1): 3053 (w), 2940 (m), 2913 (m), 2869 (m),
2828 (m), 1594 (w), 1573 (w), 1482 (w), 1436 (m), 1116 (ss),
1094 (ss), 1068 (s), 1011 (ss), 810 (m), 748 (m), 695 (ss), 528
(s), 510 (s).
Analysis (%)—found: C 65.54, H 7.43.
Acetalation degree: 77% (1H NMR), 55% (analysis).
Molar mass of the monomer unit: 319 g/mol (1H NMR).
SynthesisofP1g:748 mg(2.2 mmol)of1cand192 mgPVAin
30 ml of NMP. Extraction with boiling water for 3 days, followed
by extraction with methanol for 12 h and with Et2O over CaH2
for 2 days. The resulting polymer was dried for 18 h at 70 ◦C
and 1 mbar, yielding 484 mg (81%) of beige coloured P1g.
IR (KBr, cm−1): 3067 (m), 3053 (m), 2941 (m), 2913 (m),
2861 (m), 1955 (w), 1884 (w), 1816 (w), 1585 (m), 1570 (w),
1479 (m), 1433 (s), 1111 (s), 1048 (s), 1000 (s), 758 (s), 741
(ss), 693 (ss), 500 (m), 492 (s).
FIR (PE, cm−1): 285 (s), ν (Rh-Cl).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 30.9 (d, J
1
(P,Rh) = 151 Hz).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.7–2.0 (m
(br), CHn), 2.4 (s (br), CH2 (COD)), 3.1 (s (br), CH (COD)), 4.3
(s (br, sh), CHOR2, ROH), 5.6 (s (br, sh), CH (COD), CHO2),
5.8 (s (br), CHO2), 7.4–7.7 (m, Ar-H).
Molar mass of monomer unit: 631 g/mol.
Synthesis of P2f: 428 mg (1.34 mmol) P1f and 561 mg
(1.1 mmol) [Rh(COD)Cl]2 in 50 ml THF. Yield: 569 mg
(1.01 mmol, 75%) of yellow coloured P2f.
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = −16.3 (s).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.4 (s (br, sh),
CHn), 2.0 (s (br), CHn), 3.8 (s (br, sh), CHOR2, ROH), 4.0 (s
(br), CHOR2), 6.0 (s (br), CHO2), 6.4 (s (br), CHO2), 6.9–7.8
(m, Ar-H).
IR (KBr, cm−1): 3030 (w), 2954 (ss), 2943 (ss), 2916 (ss),
2870 (ss), 2830 (s), 1461 (s), 1386 (s), 1364 (s), 1120 (ss), 1034
(s), 1017 (ss), 814 (s), 655 (m), 629 (m).
FIR (PE, cm−1): 284 (s), ν (Rh-Cl).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 39.3 (d, J
1
Analysis (%)—found: C 74.02, H 6.20.
(P,Rh) = 144 Hz).
Acetalation degree: 68% (1H NMR), 67% (analysis).
Molar mass of monomer unit: 404 g/mol.
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.3 (s (br, sh),
CH3), 1.7 (s (br, sh), CH2 (COD), CHn), 2.5 (s (br, sh), PCH),
3.2 (s (br), CH (COD)), 4.2 (s (br, sh), OCHR, ROH), 5.4 (s (br),
CH (COD)), 5.6 (s (br, sh), CHO2), 7.5 (s (br), Ar-H).
Molar mass of monomer unit: 566 g/mol.
Synthesis of P1h: 828 mg (3.1 mmol) 1d and 204 mg PVA
in 30 ml NMP. Purification by extraction with water (16 h),
methanol (6 h), Et2O (containing CaH2 as drying agent, 16 h),
and pentane (16 h). After drying for 12 h at 90 ◦C in vac-
uum, 457 mg (75%) P1h were obtained as a colourless
solid.
Synthesis of P2g: 212 mg (0.52 mmol) P1g and 244 mg
(0.5 mmol) [Rh(COD)Cl]2 in 50 ml THF. Yield: 299 mg P2g
(0.46 mmol, 88%) as a yellow solid.

M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
85
3
IR (KBr, cm−1): 3051 (m), 2935 (s), 2912 (s), 2870 (s), 2830
(s), 1964 (w), 1896 (w), 1820 (w), 1479 (m), 1433 (s), 1186 (w),
1114 (s), 1091 (s), 1049 (s), 998 (s), 757 (s), 745 (ss), 695 (ss),
532 (m), 512 (m).
6H, CH3), 1.27 (pt, J (H,H) = 3J (H,P) = 7.5 Hz, 6H, CH3),
1.46 (dpsept, 2J (H,H) = 13.5 Hz, 3J (Heq,Hax) = 2.8 Hz,
3J (Heq,Heq) = 1.4 Hz, 2H, (OCH2)2CHHeq), 2.24 (pqt, 2J
(H,H) = 13.5 Hz, 3J (Hax,Hax) = 12.5 Hz, 3J (Hax,Heq) = 5.0 Hz,
FIR (PE, cm−1): 286 (s), ν (Rh-Cl).
2H, (OCH)2CHHax), 2.92 (s (br), 4H, PCH), 4.00 (ddpt, J
2
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 24.2 (s (br)).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.3–1.9 (m,
CHn), 2.4 (s (br), CH2 (COD)), 3.3 (s (br), CH (COD), CHOR2),
4.1 (s (br), CHOR2, ROH), 5.5 (s (br, sh), CH (COD), CHO2),
7.4–7.8 (m, Ar-H).
(H,H) = 10.7 Hz, 3J (Hax,Hax) = 12.5 Hz, 3J (Hax,Heq) = 2.8 Hz,
4J (Hax,Heq) = 1.5 Hz, 4H, OCHHax), 4.29 (ddpt, 2J
3
3
(H,H) = 10.7 Hz, J (Heq,Hax) = 5.0 Hz, J (Heq,Heq) = 1.4 Hz,
4J (Heq,Hax) = 1.5 Hz, 4H, OCHHeq), 5.54 (s, 2H, CHO2),
7.54–7.81 (m, 8H, Ar).
1
Molar mass of monomer unit: 651 g/mol.
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 45.7 (d, J
Synthesis of P2h: 476 mg (1.55 mmol) P1h and 806 mg
(1.6 mmol) [Rh(COD)Cl]2 in 60 ml THF. Yield: 730 mg
(1.32 mmol, 85%) yellow coloured P2h.
(P,Rh) = 124.4 Hz).
13C NMR (63 MHz, CDCl3, 300 K): δ (ppm) = 18.2 (s, 4C,
CH3), 19.3 (s, 4C, CH3), 22.2 (pt, 1J (C,P) = 2J (C,Rh) = 12.1 Hz,
4C, PCH), 26.1 (s, 2C, (OCH2)2CH2), 67.8 (s, 4C, OCH2), 101.5
(s, 2C, CHO2), 125.6–140.7 (m, 12C, Ar).
ESI-MS (m/z, %): 708 ([C33H50O5P2Rh·NH3]+, 100%), 691
(C33H50O5P2Rh+, 39%, M+ − Cl).
IR (KBr, cm−1): 3059 (w), 2936 (ss), 2913 (ss), 2871 (ss),
2830 (s), 1466 (m), 1388 (m), 1369 (m), 1129 (s), 1108 (s), 1056
(m), 1030 (m), 759 (s), 647 (w), 618 (w).
FIR (PE, cm−1): 280 (s), ν (Rh-Cl).
31P NMR (101 MHz, CDCl3, 300 K): δ (ppm) = 23.4 (s (br)).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.3 (s (br, sh),
CHn), 1.9 (s (br), CH2 (COD)), 2.4 (s (br), PCH), 3.5 (s (br), CH
(COD)), 4.4 (s (br, sh), OCHR, ROH), 5.3 (s (br), CH (COD)),
7.3–7.8 (m, Ar-H, CHO2).
C33H50ClO5P2Rh (%)—calc.: C 54.52, H 6.93, found: C
52.59, H 6.60.
3. Results and discussion
Molar mass of monomer unit: 554 g/mol.
{Carbonyl-chloro-trans-bis-[2-(para-
The transformation of the free hydroxyl groups of commer-
cially available PVA into cyclic 1,3-dioxanes substituted at the
2-, 4-, and 6-positions can be performed via the transacetalation
reaction in NMP under acidic catalysis (Scheme 1).
For this phosphino functionalisation of PVA, phosphino-
functionalised dimethylacetals are needed. The corresponding
para- or ortho-substituted benzaldehyde derivatives 1 were
synthesised by subsequent lithiation of the para- or ortho-
bromobenzaldehydedimethylacetal, followed by a reaction with
the corresponding chlorophosphine (Scheme 2).
The methanol released during the transacetalation reaction is
distilled off under reaction conditions. The resulting polymers
P1e–P1h can be characterised by spectroscopic methods. Char-
acterisation of polymers just by their spectroscopic properties,
however, is somewhat uncertain when their spectra cannot be
compared to those of suitable model compounds. For this rea-
son, phosphino-functionalised benzaldehyde derivatives were
synthesised with the aldehyde functionality transformed into a
diisopropylphosphinophenyl)-1,3-dioxane]rhodium(I)},
3f: At RT, 1.085 g (3.87 mmol) 1d was added to a solution of
376 mg (0.97 mmol) [Rh(CO)2Cl]2 in 15 ml dichloromethane.
After stirring for 30 min, the solvent was removed in vacuum.
The residue was washed two times with portions of 15 ml
of pentane. Purification proceeded via heat extraction with
pentane, yielding 1.272 g (1.75 mmol, 90%) yellow coloured
3f. Crystallisation was performed by layering a solution of 3f
in chloroform with pentane.
IR (KBr, cm−1): 3042 (w), 2964 (m), 2929 (m), 2965 (m),
1968 (ss), 1461 (w), 1382 (m), 1151 (m), 1100 (s), 1018 (m),
808 (m), 573 (m), 530 (m).
FIR (PE, cm−1): 295 (s), ν (Rh-Cl).
1H NMR (250 MHz, CDCl3, 300 K): δ (ppm) = 1.06 (pt, J
3
(H,H) = 3J (H,P) = 7.0 Hz, 6H, CH3), 1.09 (pt, 3J (H,H) = 3J
(H,P) = 6.9 Hz, 6H, CH3), 1.23 (pt, 3J (H,H) = 3J (H,P) = 7.5 Hz,
Scheme 1. Transacetalation of PVA, formation of the phosphino-functionalised polymers P1e–P1h and of the corresponding [Rh]-modified PVAs P2e–P2h ([Rh]:
[(COD)RhCl]) by complexation.

86
M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
Scheme 2. Formation of the ligands 1a–1d.
Fig. 4. 1H NMR spectrum of the polymer P1f (top) and its corresponding model
ligand 1f (bottom).
structure motifs. For the ligand 1e, for instance, the molecular
crystal structure shows the substitution pattern at the phenyl ring
of the benzaldehyde unit as well as the successful embedding
of the aldehyde functionalisation in the 1,3-dioxane ring system
(Fig. 3).
The spectroscopic data of the model ligands 1e–1h and their
corresponding modified PVAs P1e–P1h reveal a generally very
high agreement and confirm that the structure motifs of the lig-
ands 1e–1h also occur on the PVA derivatives. This is obvious
from a comparison of the 1H NMR spectra of 1f with P1f, where
the functionalisation of the PVA is reflected by resonances for
the methyl groups of the isopropyl residues at frequencies lower
than 1 ppm (Fig. 4). Furthermore, proton resonances of the poly-
mer P1f belonging to the aromatic protons (at 7.5 ppm) as well to
the acetal H atom (5.5 ppm) confirm the successful modification
of the PVA. Slight differences in the NMR spectra of P1f and its
model compound 1f can be explained by the fact that the model
is simple, as it does not reflect the different substitution patterns
at the polymer backbone in the 4,6-positions of the 1,3-dioxane
rings on the polymer P1f. However, resonances of the polymer
backbone are observed at 4.2 ppm and at 1.7 ppm as broad sig-
nals. Additional resonances at 3.3 ppm for the hydroxyl groups
Scheme 3. Formation of the ligands 1e–1h.
cyclic 1,3-dioxane as the simplest and smallest representative
part of the PVA backbone. In analogy to the acyclic derivatives,
the syntheses were performed by lithiation of the bromophenyl-
1,3-dioxanes functionalised in the 2-position. After this, they
reacted with the corresponding chlorophosphine (Scheme 3) to
the ligands 1e–1h.
As three of the ligands (1e, 1g [15], 1h [15]) crystallise,
their molecular structures could be determined (Table 3; Fig. 3;
Section 2). On this basis, the corresponding polymers P1e–P1h
were identified by comparing the spectroscopic data of the
model ligands 1e–1h that correlate with the absolutely defined
Fig. 3. Molecular structure of 1e in the crystal. Selected bond distances [pm] and angles [◦]: P1–C11 182.7(1), P1–C17 183.0(1), P1–C1 183.2(1), O1–C7 138.4(2),
O1–C10 144.6(2), O2–C7 139.4(2), O2–C8 143.8(2), H7–C7–C6–C5 178.6.

M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
87
of P1f are observed, whereas they are missing for the model
compound.
From the 1H NMR spectrum of P1f (Fig. 4), it can be
seen easily that the functionalisation of the PVA with diiso-
propylphosphinophenyl moieties reaches high degrees, as very
intense resonances caused by these moieties can be observed
1
in addition to the signals of the PVA backbone. Indeed, H
NMR spectroscopy is one suitable tool for the determination
of the acetalation degree of the polymers P1e–P1h. For P1f, for
instance, it can be calculated by integrating its aromatic proton
resonances and all other proton resonances. According to the
formula of P1f (Scheme 1), the integral of the aromatic protons
is set equal to 4 (n = 1 in Scheme 1) and the integral of all other
proton resonances has been determined to be 23.4, which has
to be equal to 21 + 4m (occupancy m for the vinyl alcohol unit
in Scheme 1). This leads to a final value of m = 0.6 that corre-
sponds to an acetalation degree of 77% for P1f. This corresponds
to a formula for P1f of C17H25O2P·0.6C2H4O with a molecu-
lar weight of this fictitious unit of 319 g/mol. With this high
degree of functionalisation, P1f possesses phosphorous contents
of 9.7%. If complexation of these phosphine units on the poly-
mer proceeded quantitatively with [(COD)RhCl] fragments, this
would lead to a polymer P2f with an Rh content of 18% in a
well-defined molecular environment, but bound to a polymeric
support.
Loading with Rh complex fragments of the PVA that were
phosphino-functionalised by acetalation proceeds in analogy
to the ligands 1 by reaction with [(COD)RhCl]2 in CH2Cl2
solution under the formation of pre-catalysts of the type
[chloro(COD)(phosphine)rhodium(I)] (Schemes 1 and 4) [17].
As the ligands 1a–1d at their metal-binding phosphine part
are very similar to the ligands 1e–1h, their Rh complexes 2a–2d
might act as suitable models for the description of the Rh envi-
Scheme 4. Formation of the complexes 2b, 2e, 2f, 2g, 2h (with [P]: 1b, 1e, 1f,
1g, 1h).
Fig. 5. 31P{ H} NMR spectra of the complex 2b (bottom), 2f (middle), and its
1
corresponding analogue P2f immobilised on PVA (top).
ronment of the Rh-modified PVAs as well. Accordingly, the
spectroscopic data related to the metal part (e.g. 31P{ H} NMR
1
data) of the complexes 2b and 2f exhibit values that are very
similar to those of the corresponding polymer P2f. All 31P reso-
nances are observed at about 40 ppm within 1 ppm (Fig. 5) and
Fig. 6. Molecular structures of 2b (left), 2e (centre), and 2f (right) in the crystal. Selected bond lengths [pm] and angles [◦]—2b: Rh1–C17 212.3(4), Rh1–C16
213.1(3), Rh1–C21 219.5(3), Rh1–C20 222.0(4), Rh1–P1 232.88(9), Rh1–Cl1 237.04(9), O1–C7 140.1(4), O2–C7 140.9(4), H7–C7–C6–C5 124.1; 2e: Rh1–C24
213.9(5), Rh1–C23 214.0(5), Rh1–C27 222.2(5), Rh1–C28 222.9(5), Rh1–P1 232.69(14), Rh1–Cl1 236.95(13), O1–C7 137.9(6), O2–C7 137.5(7), H7–C7–C4–C5
73.3; 2f: Rh1–C18 211.9(2), Rh1–C17 213.9(2), Rh1–C22 219.3(2), Rh1–C21 223.7(2), Rh1–P1 232.97(6), Rh1–Cl1 236.63(6), O1–C7 141.5(3), O2–C7 139.8(3),
H7–C7–C4–C5 121.8.

88
M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
Table 1
E and product distribution [%] after the hydroformylation of 1-octene, catalysed by the complexes 2, 3f or the Rh-modified polymer P2
Catalyst
Nonanal
2-Methyl-octanal
2-Ethyl-heptanal
2-Propyl-hexanal
Olefins/octane
Alcohols
2b
2e
2f
2g
48.7
50.8
55.5
49.0
52.8
48.8
49.1
50.8
46.1
26.1
36.6
38.1
36.1
38.7
37.0
36.3
38.4
34.1
36.9
18.2
9.6
7.7
6.3
8.9
7.5
10.1
8.1
9.8
10.8
0
4.3
2.4
1.9
2.9
2.4
4.3
4.1
4.7
5.6
0
0.8
1.0
0.3
0.5
0.3
0.6
0.4
0.6
0
0
0
0
0
0
0
0
0
0
2h
P2e
P2f
P2g
P2h
3f
0.6
55.7a
Experiments performed with 0.02 mmol catalyst/5 mmol octene in 50 ml dichloromethane at ∼65 ◦C and ∼50 bar syngas pressure for 15 h. The catalyses were
performed almost to complete conversion. Selectivity is based on GC/MS data.
a
Major part (>95%): 1-octene.
the 1J (P,Rh) coupling constants which could even be resolved
for the polymer P2f were determined to be ∼146 Hz.
solvent in the presence of the support. By the stoichiometric
ratio of the polymer to the support, the thickness of the result-
ing polymer layers on the support can be controlled. For the
SEM analysis shown in Fig. 7, an inorganic support has been
surrounded by a relatively thick [Rh]-modified PVA layer of
about 10 ␮m for better resolution. Generally, a more homoge-
neous thickness of the layers can be achieved by the formation
of layers of about 1 ␮m. The porosity of the inorganic support
is reflected well by the backscattered electrons, where the poly-
meric layer on the particle shows a homogeneous structure. BET
measurementsforthepureinorganicsupportrevealedaBETsur-
face of 430 m2/g, whereas a value of 22–25 m2/g was determined
for the particles surrounded by the polymer. This decrease can
be explained by the fact that the polymer layer on the inorganic
support shields its total surface. Consequently, the polymer does
not enter the inner core of the inorganic support, which is also
reflected by the EFM analyses. According to elemental EFM
analyses, Rh, P, and Cl are distributed homogeneously in the
polymeric layer on the surface of the support, whereas the ele-
ments only present in the inorganic core, such as Si or Al, are
placed inside the particle. Once formed, the layers are stable
and cannot be removed from the surface by extraction of P2d
Consequently, the complexes 2 and their spectroscopic prop-
erties allow for an unambiguous characterisation of the corre-
sponding Rh-modified polymers P2e–P2h. The X-ray analyses
performed on single crystals of the complexes 2b, 2e, 2f (Fig. 6;
Table 3), and 2g, 2h [15] additionally confirm the molecular
structures of the complexes and, hence, of the polymer-bound
complexes and do not show any irregularity [15,18].
According to our idea of developing a new approach to immo-
bilising catalysts suitable for hydrophobic substrates, the com-
plexes 2 as well as for their polymer-bound analogues, P2e–P2h
were tested for their catalytic selectivity in hydroformylation.
As substrate, the highly unpolar 1-octene was chosen,
because a solution still has to be found for the biphasic hydro-
formylation of long-chain olefins, because it was to be found
whether the highly functionalised PVA-based pre-catalysts
P2e–P2h really are unpolar enough to transform even unpolar
substrates, and because there are a lot of reference data available
in literature [19]. Other substrates or even mixtures of substrates
have not yet been analysed for discrimination as a function of
the polarity of the polymer, but such investigations certainly will
be useful.
In the hydroformylation of octene, selectivities for the com-
plexes 2 or the [Rh]-modified PVAs P2 are very similar to each
other (Table 1). The overall aldehyde selectivity is high and the
complexes do not show any high activity in the hydrogenation of
octene or the aldehydes produced. All of the pre-catalysts exhibit
isomerisation activity, transforming 1-octene into 2-octene, 3-
octene, and 4-octene, which then may also be hydroformylated
by the catalysts. As the hydroformylation activity of the termi-
nal olefin is much higher than that of the inner olefins, the most
important products representing 90% of the selectivity are n-
nonanal and 2-methyloctanal, their formation ratio being 1.25:1.
This low n/iso ratio is typical of hydroformylation reactions in
the absence of a large phosphine excess [19].
The polymeric pre-catalysts may be adsorbed on an inor-
ganic support. Especially when the functionalisation of PVA has
reached a high degree, the resulting polymers can be dissolved
easily in organic solvents, such as dichloromethane. Adsorption
on the inorganic supports proceeds by a simple removal of the
˚
Fig. 7. Surface of an inorganic support (molecular sieve, 5 A, particle size
2.4–4.8 mm) surrounded by a layer of a Rh-loaded phosphino-functionalised
PVA.

M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
89
Table 2
E and product distribution [%] in the hydroformylation of 1-octene, during 10 consecutive runs, catalysed by the Rh-modfied PVA P2f
Run
Nonanal
2-Methyl-octanal
2-Ethyl-heptanal
2-Propyl-hexanal
Olefins/octane
Alcohols
1
2
3
4
5
6
7
8
9
50.4
50.3
49.7
48.8
49.2
48.4
49.8
48.3
48.5
49.2
36.0
35.5
34.8
33.6
35.2
34.0
35.3
32.2
31.7
31.7
9.7
9.9
8.2
9.8
10.1
8.5
8.5
7.6
5.0
6.3
3.3
3.7
3.1
4.0
4.4
3.5
3.8
3.3
0.9
3.2
0.6
0.6
0.8
1.0
1.1
5.2
2.5
5.1
13.9
9.6
0
0
3.4
2.8
0
0.4
0
3.4
0
0
10
Experiments performed with 0.02 mmol Rh-catalyst/5.2 mmol octene in 50 ml pentane at ∼65 ◦C and ∼60 bar syngas pressure for 15 h. Constitution of the mixtures
based on GC/MS data.
with THF or methanol for 3 days. It may therefore be expected
that the layers will also be stable under catalytic conditions and
suitable for an application in interfacial catalysis.
catalysis. Alternatively, the phosphino-functionalised polymer
P1f (not containing any Rh) was subjected to a reaction with
[(CO)2RhCl]2 and the IR spectrum of the resulting Rh-modified
polymer was measured (Fig. 8, bottom). The high congruence
of this IR spectrum with the one of P2f after catalysis shows that
these two polymers have an identical composition. The existence
of the two other CO stretching vibrations at 2087 and 2004 cm−1
can only be explained by the presence of complexes on the poly-
mer P2f of the type [cis-(dicarbonyl)(chloro)(phosphine)Rh(I)]
with two CO ligands coordinated to the Rh centre, one in trans
position to the phosphine and the other to the chloro ligand. This
coordination behaviour leads to two CO stretching vibrations
that are significantly separated in the IR spectrum and in high
agreement with literature values [20]. These findings plus the
result of a single-crystal X-ray analysis of crystals of 3f (Fig. 9;
Table 3) confirm that the catalyst rest state on the polymer P2f
after 10 uses in catalysis is {cis-(dicarbonyl)(chloro)[2-(para-
diisopropylphosphinophenyl)-1,3-dioxane]Rh(I)}.
In order to demonstrate that the Rh-modified PVA can act
as a recyclable catalyst, the polymer P2f was selected as a
pre-catalyst in consecutive experiments. This choice was based
on the well-resolved NMR spectra obtained from this polymer
(Fig. 5). As the spectroscopic characterisation of the polymer
itself, once it has been adsorbed on an inorganic surface, is more
complicated than for an unsupported polymer, these experiments
were performed with P2f embedded in a Soxhlet-extraction
container. In order to reduce the solubility of the polymer in
the system, pentane instead of dichloromethane was used as
a co-solvent in these experiments. Ten consecutive runs for
the hydroformylation of 1-octene were performed. This cor-
responded to an overall time of more than 150 h of use in
catalysis, with a certain decrease in activity, however (Table 2).
This is obvious from run no. 10, where nearly 10% of the 1-
octene was still found to be present as unreacted substrate.
This decrease in activity may be caused by several factors: (i)
uncontrolled leaching of the Rh into the organic phase without
any break of the acetals on the polymer, (ii) an uncontrolled
cleavage of the acetals of the polymer, (iii) a certain oxida-
tion of the ligand, then leading to metal leaching, or (iv) the
formation of less active complexes on the polymer with possi-
ble partial leaching of the metal. All these cases will have an
influence on the spectroscopic data of the polymer. Therefore,
the polymeric catalyst P2f was analysed by IR spectroscopy
after 10 consecutive uses in the hydroformylation of 1-octene
(Fig. 8). In the region of coordinated CO stretching vibrations,
the IR spectrum shows three absorptions at 2087, 2004, and
1972 cm−1, respectively (Fig. 8, middle). From a catalytic run
with the model complex 2f, a complex of the type [carbonyl-
chloro-trans-(disphosphine)rhodium(I)] (3f) was isolated, with
the phosphine being 2-(para-diisopropylphosphinophenyl)-1,3-
dioxane, 1f. However, 3f can also be synthesised by the reaction
of 1f with [Rh(CO)2Cl]2. The IR spectrum of 3f shows one CO
absorption band at 1968 cm−1 (Fig. 8, top) that corresponds well
Concerning the quantification of the amount of catalyti-
1
cally active species, 31P{ H} NMR spectroscopic investiga-
tions were carried out on the polymer obtained by the reaction
of [(CO)2RhCl]2 with P1f. Two 31P resonances were observed,
1
one at 45.7 ppm and the other at 46.6 ppm with a JPRh
coupling constant of 126 Hz, respectively. By comparison
to that of the polymeric P2f after catalysis at 1972 cm−1
.
Fig. 8. IR spectra of {carbonyl-chloro-trans-bis-[2-(para-diisopropylphos-
phinophenyl)-1,3-dioxane]rhodium(I)} (top), P2f after 10 runs in the hydro-
formylation of 1-octene (middle), and an alternatively synthesised Rh-loaded
polymer obtained from the reaction of P1f with [Rh(CO)2Cl]2 (bottom).
In the IR spectrum of P2f after catalysis, an absorption at
1724 cm−1 was observed. It was caused by traces of aldehy-
des, which were present in the polymer matrix after its use in

90
M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
Fig. 9. Molecular structure of 3f in the crystal. 3f crystallises in the centrosymmetric triclinic space group P-1 with the Rh atom located at the inversion centre.
Selected bond lengths [pm] and angles [◦]: Rh1–C17A 173.2(7), Rh1–P1 232.76(5), Rh1–Cl1 241.9(3), C17A–O3A 116.2(7) P1–Rh1–P1A 180.0 (0), Cl1–Rh1–C17A
179.2(2), Cl1–Rh1–P1 90.12(7), P1–Rh1–C17A 89.8(2).
Table 3
Crystallographic data
Compound
1e
2b
2e
2f
3f
Empirical formula
Formula weight
Crystal size
C
22H21O2P
C23H37ClO2PRh
514.86
C30H33ClO2PRh
594.89
0.3 mm × 0.2 mm
× 0.15 mm
Triclinic
C24H37ClO2PRh
526.87
C33H50ClO5P2Rh
727.06
348.36
0.35 mm × 0.35 mm
× 0.4 mm
Monoclinic
P21/c (no. 14)
0.40 mm × 0.40 mm
0.4 mm × 0.35 mm
0.4 mm × 0.35 mm
× 0.20 mm
× 0.2 mm
× 0.2 mm
Crystal system
Space group
Monoclinic
Monoclinic
Triclinic
Pn (no. 7)
P-1 (no. 2)
P21/c (no. 14)
P-1 (no. 2)
Unit cell dimensions
a = 1044.3(1)
b = 1912.8(2)
c = 935.4(1) pm
α = 90.0◦
β = 97.076(2)◦
γ = 90.0◦
a = 860.12(3)
b = 807.35(3)
c = 1738.80(6) pm
α = 90.0◦
β = 102.041(1)◦
γ = 90.0◦
a = 808.64(14)
b = 1243.8(2)
c = 1347.2(2) pm
α = 88.447(3)◦
β = 79.137(3)◦
γ = 85.682(3)◦
a = 823.2(1)
b = 1402.1(2)
c = 2100.0(3) pm
α = 90.0◦
β = 92.334(2)◦
γ = 90.0◦
a = 857.33(7)
b = 937.70(8)
c = 11.4291(9) pm
α = 107.327(1)◦
β = 93.938(1)◦
γ = 95.974(1)◦
Volume
Z
Density (calculated)
Temperature
θ range
1854.5(4) × 106 pm3
1180.89(7) × 106 pm3
1326.8(4) × 106 pm3
2421.9(5) × 106 pm3
867.5(1) × 106 pm3
4
2
2
4
1
1.248 g/cm3
200 K
1.448 g/cm3
200 K
1.489 g/cm3
200 K
1.445 g/cm3
200 K
1.392 g/cm3
200 K
1.52 ≤ θ ≤ 28.28◦
ω-Scan, ꢀω = 0.3◦
−13 ≤ h ≤ 13,
−25 ≤ k ≤ 25,
−12 ≤ l ≤ 12
1.68 ≤ θ ≤ 28.27◦
ω-Scan, ꢀω = 0.3◦
−11 ≤ h ≤ 11,
−10 ≤ k ≤ 10,
−23 ≤ l ≤ 22
1.54 ≤ θ ≤ 28.26◦
ω-Scan, ꢀω = .45◦
−10 ≤ h ≤ 10,
−16 ≤ k ≤ 16,
−17 ≤ l ≤ 17
1.75 ≤ θ ≤ 28.27◦
ω-Scan, ꢀω = .45◦
−10 ≤ h ≤ 10,
−18 ≤ k ≤ 18,
−27 ≤ l ≤ 27
1.88 ≤ ␪ ≤ 28.31◦
ω-Scan, ꢀω = 0.3◦
−11 ≤ h ≤ 11,
−12 ≤ k ≤ 12,
−14 ≤ l ≤ 14
Scan
Index ranges
Number of
Reflections measured
Unique reflections
Reflections observed
18798
4506
3784 (I > 2σ)
12151
5419
4894 (I > 2σ)
14456
6332
3538 (I > 2σ)
25369
5904
4771 (I > 2σ)
9207
4134
3574 (I > 2σ)
Parameters refined
Residual electron density
Corrections
Structure solution
Structure refinement
Programs used
231
271
320
313
0.574 × 10−6 e/pm3
0.412 × 10−6 e/pm3
2.139 × 10−6 e/pm3
0.513 × 10−6 e/pm3
0.541 × 10−6 e/pm3
Lorentz and polarisation, exp. absorption correction [21]
Direct methods
Full matrix least square on F2
SHELX-97 [22], xpma [23], winray [24]
R indices
R1 = 0.0422
(I > 2σ)
R1 = 0.0319
(I > 2σ)
R1 = 0.0567
(I > 2σ)
R1 = 0.0315
(I > 2σ)
R1 = 0.0324 (I > 2σ)
Rw = 0.1244 (all
data on F2)
Rw = 0.0646 (all
data on F2)
Rw = 0.1362 (all
data on F2)
Rw = 0.0710 (all
data on F2)
Rw = 0.0769 (all data
on F2)
1
with the 31P NMR spectrum of 3f having its resonance at
45.7 ppm and a coupling constant of 124.4 Hz, the second res-
onance can be assigned to the polymer bound analogous to
3f that exhibits hardly any catalytic activity. The resonance at
45.7 ppm with its JPRh coupling constant of 143 Hz is there-
fore caused by the catalytically active species of the type [cis-
(dicarbonyl)(chloro)(phosphine)Rh(I)]. Integration of these two
signals yields a 2:1 ratio for the cis-(dicarbonyl) complex as

M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
91
compared to the minor complex on the polymer of the type
[(carbonyl)(chloro)-trans-(disphosphine)rhodium(I)]. As in this
complex, two phosphine ligands are bound to one Rh atom, the
Rh distribution between the complexes resulting in 4:1, leav-
ing 80% of the polymer-bound Rh in the catalytically active
species [cis-(dicarbonyl)(chloro)(phosphine)Rh(I)]. This corre-
sponds to 66% of the initial Rh content on the polymer and
explains the reduced activity of the system. However, it is
implied that at least 17% of the initial Rh contents have to be
removed from the polymer by mobilisation. Indeed, crystals of
[(COD)RhCl]2 were isolated from the yellow coloured solu-
tions of the first two catalytic runs. In the later runs, however, no
coloured solutions were obtained. ICP analyses of the solutions
of all runs were not performed. For a future improvement of the
system, they will have to be included. Nevertheless, our results
show clearly that after 10 consecutive uses (or more than 150 h)
in the hydroformylation, the polymeric pre-catalyst P2f has
changedbytheformationofwell-definedcomplexesonthepoly-
mer. No uncontrolled leaching or degradation of the polymer-
bound complexes was observed. A certain loss of Rh was noticed
at the beginning, but once the stable complexes were formed on
the polymer, it could not be detected visually any longer. Due to
these results, it may well be expected that the modified PVA sys-
tems might be an alternative to existing technologies for catalyst
immobilisation in the future even for a longer use.
hence, mobile [(COD)RhCl]2. These first attempts of using
phosphino-functionalised PVA for the immobilisation of transi-
tion metal catalysts have shown that this system has its potential
and might represent an alternative for catalyst immobilisation.
Acknowledgements
We are very grateful to Mr. G. Zwick for the measurements of
the MS data. Furthermore, we wish to thank Forschungszentrum
Karlsruhe for the financial and Prof. Eckhard Dinjus for the
general support of this work.
References
[1] P.T. Anastas, J.C. Warner, Green Chemistry: Theory and Practice, Oxford
University Press, Oxford, 1998.
[2] U. Mu¨ller-Eisen, H. Hulpke, Ullmann’s Encyclopedia of Industrial
Chemistry, Sustainability and the Chemical Industry—Ethical Aspects
and Approaches, Wiley-VCH, 2002, doi:10.1002/14356007.b07 001.
[3] L. Jourdan, F. Jonckheere, U. Geffarth, J. Busson, Ullmann’s Encyclope-
dia of Industrial Chemistry, Environmental Management in the Chemical
Industry, Wiley-VCH, 2002, doi:10.1002/14356007.b07 359.
[4] D. Vogt, in: B. Cornils, W.A. Herrmann (Eds.), Aqueous-Phase
Organometallic Catalysis, Wiley-VCH, Weinheim, 1998, p. 541 (Chapter
7.1, and references cited therein).
[5] B. Cornils, E.G. Kuntz, in: B. Cornils, W.A. Herrmann (Eds.), Aqueous-
Phase Organometallic Catalysis, Wiley-VCH, Weinheim, 1998, p. 270
(Chapter 6.1, and references cited therein).
4. Conclusion
[6] M.L. Hallensleben, Ullmann’s Encyclopedia of Industrial Chemistry,
Poly(Vinyl Acetals), Wiley-VCH, 2002, doi:10.1002/14356007.a21 743.
[7] (a) T. Malmstro¨m, C. Andersson, J. Hjortkjaer, J. Mol. Catal. A: Chem.
139 (1999) 139;
It was demonstrated that a phosphino functionalisation
of polyvinyl alcohol (PVA) can be achieved. The phosphino
substituents were introduced in the polymer by a one-step
transacetalation reaction. High acetalation degrees were
reached. Polymers with Rh contents higher than 15% and
the transition metal in a defined molecular environment were
obtained. The phosphino-modified PVAs or the Rh-modified
PVAs resulting from the complexation of the phosphino groups
towards a [(COD)RhCl] complex were characterised unam-
biguously by comparison of their spectroscopic data with those
of model compounds, the syntheses of which were reported as
well. Finally, the Rh-modified PVAs were tested in the catalytic
hydroformylation of 1-octene. Their selectivity in the product
distribution was found to be identical to that of the model
complexes. The Rh-modified PVA was adsorbed on an inorganic
support under the formation of stable polymer layers on its
surface, which cannot be mobilised by extraction with organic
solvents. The stability of these layers, together with their tune-
able polarity, allow for an application in interfacial catalysis.
Analyses performed on a polymer that had not been adsorbed on
an inorganic support and had been used 10 times as the catalyst
in the hydroformylation of 1-octene revealed that the polymeric
pre-catalyst P2f had changed by the formation of well-defined
complexes on the polymer. Sixty-six percent of the initial Rh
contents were transformed in the catalytically active polymer-
bound species [cis-(dicarbonyl)(chloro)(phosphine)Rh(I)],
17% were found to rest in a catalytically inactive form of
polymer-bound [carbonyl-chloro-trans-(diphosphine)Rh(I)],
and 17% were lost in the first runs in the form of soluble and,
(b) Z. Li, Q. Peng, Y. Yuan, Appl. Catal. A: Gen. 239 (2003) 79;
(c) U.J. Ja´uregui-Haza, M. Dessoudeix, Ph. Kalck, A.M. Wilhelm, H.
Delmas, Catal. Today 66 (2001) 297;
(d) M. Benaissa, U.J. Ja´uregui-Haza, I. Nikov, A.M. Wilhelm, H. Del-
mas, Catal. Today 79–80 (2003) 419;
(e) Ph. Kalck, M. Dessoudeix, Coord. Chem. Rev. 190–192 (1999)
1185.
[8] (a) C.P. Mehnert, R.A. Cook, N.C. Dispenziere, M. Afeworki, J. Am.
Chem. Soc. 124 (2002) 12932;
(b) A. Riisager, P. Wasserscheid, R. van Hal, R. Fehrmann, J. Catal. 219
(2003) 452;
(c) A. Riisager, K.M. Eriksen, P. Wasserscheid, R. Fehrmann, Catal.
Lett. 90 (2003) 149;
(d) A. Riisager, R. Fehrmann, S. Flicker, R. van Hal, M. Haumann, P.
Wasserscheid, Angew. Chem. Int. Ed. Eng. 44 (2005) 815.
[9] (a) G. Oehme, I. Grassert, E. Paetzold, R. Meisel, K. Drexler, H.
Fuhrmann, Coord. Chem. Rev. 185–186 (1999) 585;
(b) T.J. Dickerson, N.N. Reed, K.D. Janda, Chem. Rev. 102 (2002)
3325;
(c) C.E. Song, S. Lee, Chem. Rev. 102 (2002) 3495;
(d) Z. Lu, E. Lindner, H.A. Mayer, Chem. Rev. 102 (2002) 3543;
(e) R. van Heerbeek, P.C.J. Kamer, P.W.N.M. van Leeuwen, J.N.H. Reek,
Chem. Rev. 102 (2002) 3717;
(f) P.W.N.M. van Leeuwen, A.J. Sandee, J.N.H. Reek, C.J. Paul, Kamer,
J. Mol. Catal. A: Chem. 182–183 (2002) 107.
[10] G. Giordano, R.H. Crabtree, Inorg. Synth. 19 (1979) 218.
[11] J.A. Mc Cleverty, G. Wilkinson, Inorg. Synth. 8 (1966) 211.
[12] (a) J.E. Hoots, T.B. Rauchfuss, D.A. Wrobleski, Inorg. Synth. 21 (1981)
175;
(b) G.P. Schiemenz, H. Kaack, Liebigs Ann. Chem. (1973) 1480.
[13] V.D. Bianco, S. Doronzo, Inorg. Synth. 16 (1976) 161–163.
[14] G.D. Vaughn, C.E. Strouse, J.A. Gladysz, J. Am. Chem. Soc. 108 (1986)
1462–1473.

92
M. Ahlmann et al. / Journal of Molecular Catalysis A: Chemical 249 (2006) 80–92
[15] M. Ahlmann, O. Walter, J. Organomet. Chem. 689 (2004) 3117.
[16] F.H. Allen, O. Kennard, Chem. Des. Autom. News 8 (1993) 31.
[17] J. Chatt, L.M. Venanzi, J. Chem. Soc. (1957) 4735.
[18] (a) B.D. Murray, H. Hope, J. Hvoslef, P.P. Power, Organometallics 3
(1984) 657;
(b) C.D. Frohning, C.W. Kohlpaintner, H.-W. Bohnen, in: B.
Cornils, W.A. Herrmann (Eds.), Applied Homogeneous Catalysis with
Organometallic Compounds, vol. 1, Wiley, 2003, p. 31;
(c) F. Plourde, K. Gilbert, J. Gagnon, P.D. Harvey, Organometallics 22
(2003) 2862;
(b) M. Iglesias, C. Del Pino, A. Corma, S. Garcia-Blanco, S. Martinez-
Carrera, Inorg. Chim. Acta 127 (1987) 215;
(d) A. Solsona, J. Suades, R. Mathieu, J. Organomet. Chem. 669 (2003)
172;
(c) O.J. Scherer, M. Florchinger, K. Gobel, J. Kaub, W.S. Sheldrick,
Chem. Ber. 121 (1988) 1265;
(e) E. Paetzold, G. Oehme, C. Fischer, M. Frank, J. Mol. Catal. A 200
(2003) 95;
(d) A.S.C. Chan, C.-C. Chen, R. Cao, M.R. Lee, S.-M. Peng, G.H. Lee,
Organometallics 16 (1997) 3469;
(f) C. Liu, J. Jiang, Y. Wang, F. Cheng, Z. Jin, J. Mol. Catal. A 198
(2003) 23;
(e) A.M.Z. Slawin, M.B. Smith, J.D. Woollins, J. Chem. Soc. Dalton
Trans. (1996) 1283;
(f) Q.L. Horn, D.S. Jones, R.N. Evans, C.A. Ogle, T.C. Master-
(g) D.E. Bryant, M. Kilner, J. Mol. Catal. A 193 (2003) 83.
[20] G. Canepa, C.D. Brandt, H. Werner, Organometallics 23 (2004) 1140.
[21] SADABS, Siemens Area Detector Absorption Correction Programme,
Siemens, 1997.
[22] SHELX-97, G.M. Sheldrick, Universita¨t Go¨ttingen, 1997.
[23] xpma, L. Zsolnai, G. Huttner, Universita¨t Heidelberg, 1994.
[24] Winray, R. Soltek, Universita¨t Heidelberg, 1997.
man, Acta Crystallogr., Sect.
m51.
E (Struct. Rep. Online) 58 (2002)
[19] (a) I. Horvath (Ed.), Encyclopedia of Catalysis, vol. 3, Wiley (2003)
773;