Influence of the Bite Angle on the Hydroformylation of Internal Olefins to Linear Aldehydes

Article abstract of DOI:10.1021/om034012f

The effect of bite angle on the hydroformylation of internal olefins to linear aldehydes was analyzed. A novel series of ligands that induced the wide bite angles were synthesized. The high-pressure infrared and high-pressure nuclear magnetic resonance (NMR) studies showed that the ligands adopted a bis-equatorial mode in the trigonal bipyramidal rhodium complex. The results show that the selectivity-determining step remain unchanged between the terminal and internal olefins.

Full text of DOI:10.1021/om034012f

5358
Organometallics 2003, 22, 5358-5369
In flu en ce of th e Bite An gle on th e Hyd r ofor m yla tion of
In ter n a l Olefin s to Lin ea r Ald eh yd es
Raymond P. J . Bronger, Paul C. J . Kamer, and Piet W. N. M. van Leeuwen*
Institute of Molecular Chemistry, University of Amsterdam, Nieuwe Achtergracht 166,
1018 WV Amsterdam, The Netherlands
Received J uly 2, 2003
A novel series of ligands (1-10) that induce wide bite angles (106° < ân < 131°) has been
synthesized. Compared to the xantphos series (e.g., 13 versus 5), the introduction of the
phosphacyclic moiety results in ligands with a slightly larger bite angle. High-pressure IR
and high-pressure NMR studies of the (diphosphine)RhH(CO)2 complexes show that most
ligands (3-7) adopt a bis-equatorial binding mode exclusively in the trigonal bipyramidal
rhodium complex. Subtle changes in ligand structure have a large impact on activity and
selectivity in the hydroformylation of 1-octene and trans-2-octene. Rates up to 3275 (mol
aldehyde)(mol Rh)-1 h-1 (p(CO/H2) ) 20 bar, T ) 353 K, [Rh] ) 1 mM, [1-octene] ) 637
mM) and regioselectivities > 99% toward the linear product were obtained when 1-octene
was used as substrate. For trans-2-octene rates up to 250 (mol aldehyde)(mol Rh)-1 h-1 (p(CO/
H2) ) 3.6 bar, T ) 393 K, [Rh] ) 1 mM, [trans-2-octene] ) 640 mM) and high regioselectivities
up to 96% toward the linear product, which are unprecedented, were obtained. A correlation
between the selectivity for the hydroformylation of 1-octene and trans-2-octene has been
observed, suggesting that the selectivity-determining step remains unchanged between
terminal and internal olefins. Ligands with a larger bite angle lead to more selective systems,
but above 125° the regioselectivity drops. Furthermore it is no longer the selective formation
of linear alkyl species that determines the high regioselectivities. Instead the differences in
rate of â-hydrogen elimination from the branched alkyl intermediate and the linear alkyl
intermediate versus CO insertion determine the regioselectivity. For both substrates a
decrease in activity with an increase in bite angle is observed. It is suggested that the
aforementioned rates of â-hydrogen elimination versus CO insertion must play a crucial
role in this bite angle effect on activity, because previous studies have shown that an increase
in bite angle leads to an increase in activity.
In tr od u ction
angle θ and the electronic parameter ø to classify
phosphorus ligands with respect to their steric bulk and
phosphine basicity.12 Both parameters have been used
extensively as a measure of ligand properties in hydro-
formylation studies.13,14 Casey and Whiteker developed
the concept of natural bite angle (the P-M-P angle)
as an additional characteristic of diphosphine ligands
to rationalize the correlation found between the regio-
selectivity and the coordination mode of bidentate
ligands possessing wide bite angles.5 Many articles and
reviews have appeared that discuss the effect of the bite
angle on activity and selectivity for a number of catalytic
reactions.15-20 The bite angle induces not only steric
effects but also electronic effects, as the P-M-P angles
To date, the hydroformylation reaction is one of the
most important homogeneously catalyzed reactions in
the world, covering an annual production of almost eight
million tons of aldehydes and alcohols, from which
butanal and 2-ethylhexanol are the most important
products.1,2 Nowadays, the design of new ligands and
systematic tuning of their electronic and steric proper-
ties are common research topics in this area, as it
provides insight in regio- and chemoselectivity and the
overall catalytic activity.3-11 Tolman introduced the cone
* Corresponding author. E-mail: pwnm@science.uva.nl.
(1) Weissermel, K.; Arpe, H.-J Industrielle Organische Chemie; VCH
Verlagsgesellschaft mbH: Weinheim, 1988.
(8) Klein, H.; J ackstell, R.; Wiese, K.-D.; Borgmann, C.; Beller, M.
Angew. Chem., Int. Ed. Commun. 2001, 40, 3408-3411.
(9) Selent, D.; Hess, D.; Wiese, K. D.; Rottger, D.; Kunze, C.; Borner,
A. Angew. Chem., Int. Ed. 2001, 40, 1696-1698.
(10) van der Veen, L. A.; Kamer, P. C. J .; van Leeuwen, P. W. N.
M. Organometallics 1999, 18, 4765-4777.
(2) Rhodium Catalyzed Hydroformylation; van Leeuwen, P. W. N.
M., Claver, C., Eds.; Kluwer Academics Publishers: Dordrecht, 2000.
(3) van der Veen, L. A.; Boele, M. D. K.; Bregman, F. R.; Kamer, P.
C. J .; van Leeuwen, P. W. N. M.; Goubitz, K.; Fraanje, J .; Schenk, H.;
Bo, C. J . Am. Chem. Soc. 1998, 120, 11616-11626.
(4) van der Veen, L. A.; Keeven, P. H.; Schoemaker, G. C.; Reek, J .
N. H.; Kamer, P. C. J .; van Leeuwen, P. W. N. M.; Lutz, M.; Spek, A.
L. Organometallics 2000, 19, 872-883.
(5) Casey, C. P.; Whiteker, G. T. Isr. J . Chem. 1990, 30, 299-304.
(6) Casey, C. P.; Paulsen, E. L.; Beuttenmueller, E. W.; Proft, B.
R.; Petrovich, L. M.; Matter, B. A.; Powell, D. R. J . Am. Chem. Soc.
1997, 119, 11817-11825.
(7) Casey, C. P.; Paulsen, E. L.; Beuttenmueller, E. W.; Proft, B.
R.; Matter, B. A.; Powell, D. R. J . Am. Chem. Soc. 1999, 121, 63-70.
(11) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J .; van
Leeuwen, P. W. N. M. Organometallics 1995, 14, 3081.
(12) Tolman, C. A. Chem. Rev. 1977, 77, 313-348.
(13) Moser, W. R.; Papile, C. J .; Brannon, D. A.; Duwell, R. A.;
Weininger, S. J . J . Mol. Catal. 1987, 41, 271-292.
(14) van Leeuwen, P. W. N. M.; Roobeek, C. F. J . Organomet. Chem.
1983, 258, 343-350.
(15) Dierkes, P.; van Leeuwen, P. W. N. M. J . Chem. Soc., Dalton
Trans. 1999, 1519-1529.
10.1021/om034012f CCC: $25.00 © 2003 American Chemical Society
Publication on Web 11/11/2003

Hydroformylation of Internal Olefins
Organometallics, Vol. 22, No. 25, 2003 5359
Nevertheless, catalysts that fulfill all these require-
ments are not per se good catalysts for the hydroformy-
lation of internal olefins to linear aldehydes.28 Although
considerable progress in this field has been made by the
use of bulky phosphites22,24,29 and phosphonites,25 these
systems suffer from a low long-term stability due to
alcoholysis, hydrolysis, and thermal instability.30,31
Phosphines are more resistant to these degradation
reactions, but afford in general less active catalysts. To
date, only few rhodium-diphosphine systems are known
that show a respectable activity and selectivity for the
hydroformylation of internal olefins. Van Leeuwen et
al. introduced phenoxaphosphino- and dibenzophos-
phole-modified xanthene ligands (11 and 12, Figure
2),10,21 while Beller et al. introduced Naphos-type ligands
with various electron-withdrawing substituents (Figure
3).8
Recently, the effect of natural bite angle on the
catalytic activity and selectivity on the hydroformylation
of terminal olefins was investigated, for which a series
of diphosphine ligands with different xanthene-type
backbones was prepared.4,11 These ligands, however, are
not suitable for a study of the bite angle effect on the
hydroformylation of internal alkenes to linear alde-
hydes, as the rate of isomerization from the internal
olefin to the terminal olefin is too low to ensure fast and
selective catalysis. Xanthene-type ligands bearing phos-
phacyclic moieties do give systems that are suitable for
the hydroformylation of internal alkenes.10,21 To obtain
a better understanding of how the bite angle affects the
hydroformylation activity and selectivity of internal
olefins, we synthesized a range of phenoxaphosphino-
modified xanthene-type ligands that show a regular
increase in natural bite angle (Figure 3, ligands 1-8).
In addition, two new dibenzophosphole xanthene-type
ligands have been synthesized (9 and 10) to obtain a
better insight into the effect of ligand backbone and
other phosphacyclic moieties. The coordination chem-
istry and the performance in the rhodium-catalyzed
hydroformylation of 1-octene and trans-2-octene for this
novel series of ligands will be described. Furthermore,
deuterioformylation reactions of 1-hexene with ligands
1, 5, and 7 were performed. The number and position
of the deuterium atoms that are incorporated in the
aldehyde and isomerized hexenes give detailed informa-
tion about the reversibility of the different reaction steps
during hydroformylation.28,32-35
F igu r e 1. ee-ae equilibrium in rhodium trigonal bipyra-
midal complexes.
clearly affect the nature of the complex or intermediate
states. In a review by Freixa and van Leeuwen an
attempt is made to separate the contributions of the
steric bite angle effect and the electronic bite angle effect
to the selectivities and rates for several catalytic reac-
tions, including hydroformylation.16 For hydroformyla-
tion it was first believed that the coordination mode of
the diphosphine in the (diphosphine)RhH(CO)2 complex,
either equatorial-equatorial (ee) or apical-equatorial
(ae) (Figure 1), was a key factor in controlling the
regioselectivity of the hydroformylation reaction. Later
studies by van Leeuwen et al.3 showed that the coordi-
nation mode is not the sole factor determining the
outcome of the reaction.
It has been reported that phosphacyclic derivatives
of different ligands with the right backbone, e.g., ligands
based on 2,2′-dimethylbisphenyl, and xanthene back-
bones, like cyclic phosphines,10,21 cyclic phosphites,9,22-24
cyclic phosphonites,25 and cyclic phosphorus diamides,26
result in catalysts that show a high activity and
selectivity for the hydroformylation of terminal olefins.
Especially the catalytic activity is enhanced consider-
ably compared to noncyclic analogues. Decreased phos-
phorus basicity and increased electron “elasticity”,
which is attributed to the extended conjugation of the
ligand π-system, are considered to be the main reasons
for this observed effect.4,10,13,27
Generally terminal olefins are used as substrates,
while from an economical point of view internal olefins
or a mixture of internal olefins and terminal olefins are
substrates of choice, yet linear aldehydes and alcohols
are the preferred products. To obtain linear aldehydes
from internal olefins, the system has to catalyze isomer-
ization between the internal and terminal olefin, as the
thermodynamic mixture contains in general less than
5% of the terminal olefin. In addition the hydroformy-
lation of the terminal olefin must be many times faster
than the hydroformylation of the internal olefin. Also,
a high selectivity toward the linear aldehyde is required.
(16) Freixa, Z.; van Leeuwen, P. W. N. M. J . Chem. Soc., Dalton
Trans. 2003, 1890-1901.
Resu lts
(17) Kamer, P. C. J .; Reek, J . N. H.; van Leeuwen, P. W. N. M.
Chemtech 1998, 27-33.
Syn th esis. A total of 10 novel ligands that are closely
related in either backbone structures or phosphine
(18) Kamer, P. C. J .; van Leeuwen, P. W. N. M.; Reek, J . N. H. Acc.
Chem. Res. 2001, 34, 895.
(19) van Leeuwen, P. W. N. M.; Kamer, P. C. J .; Reek, J . N. H.;
Dierkes, P. Chem. Rev. 2000, 100, 2741.
(20) van Leeuwen, P. W. N. M.; Kamer, P. C. J .; van der Veen, L.
A.; Reek, J . N. H. Chin. J . Chem. 2001, 19, 1-8.
(21) van der Veen, L. A.; Kamer, P. C. J .; van Leeuwen, P. W. N.
M. Angew. Chem., Int. Ed. 1999, 38, 336-338.
(22) Billig, E.; Abatjoglou, A. G.; Bryant, D. R.; Murray, R. E.;
Maher, J . M. (Union Carbide Corp.) Eur. Pat. Appl. EP 213,639, 1987.
(23) Billig, E.; Abatjoglou, A. G.; Bryant, D. R.; Murray, R. E.;
Maher, J . M. U.S. Pat. US 4,717,775, 1988.
(28) van der Slot, S. C.; Duran, J .; Luten, J .; Kamer, P. C. J .; van
Leeuwen, P. W. N. M. Organometallics 2002, 21, 3873-3883.
(29) Burke, P. M.; Garner, J . M.; Tam, W.; Kreutzer, K. A.;
Teunissen, A. J . J . M.; Snijder, C. S.; Hansen, C. B. (DSM/Du Pont de
Nemours) WO97/33854, 1997.
(30) Kamer, P. C. J .; Reek, J . N. H.; van Leeuwen, P. W. N. M.
Rhodium Phosphite Catalysts. In Rhodium Catalyzed Hydroformyla-
tion; van Leeuwen, P. W. N. M., Claver, C., Eds.; Kluwer Academic
Publishers: Dordrecht, 2000; pp 35-62.
(31) van Leeuwen, P. W. N. M. Appl. Catal. A 2001, 212, 61-81.
(32) Casey, C. P.; Petrovich, L. M. J . Am. Chem. Soc. 1995, 117,
6007-6014.
(24) Dieleman, C. B.; Kamer, P. C. J .; Reek, J . N. H.; van Leeuwen,
P. W. N. M. Helv. Chim. Acta 2001, 84, 3269-3279.
(25) Selent, D.; Wiese, K.-D.; Rottger, D.; Borner, A. Angew. Chem.,
Int. Ed. 2000, 39, 1639-1641.
(33) Lazzaroni, R.; Settambolo, R.; Raffaelli, A.; Pucci, S.; Vitulli,
G. J . Organomet. Chem. 1988, 339, 357-365.
(26) van der Slot, S. C.; Kamer, P. C. J .; van Leeuwen, P. W. N. M.;
Fraanje, J .; Goubitz, K.; Lutz, M.; Spek, A. L. Organometallics 2000,
19, 2504-2515.
(34) Lazzaroni, R.; Uccello-Barretta, G.; Benetti, M. Organometallics
1989, 8, 2323-2327.
(35) Consiglio, G.; Morandini, F.; Haelg, P.; Pino, P. J . Mol. Catal.
1990, 60, 363-374.
(27) Hayashi, T.; Tanaka, M.; Ogata, I. J . Mol. Catal. 2003, 6, 1-9.

5360 Organometallics, Vol. 22, No. 25, 2003
Bronger et al.
F igu r e 2. Phosphacyclic xanthene type ligands. 11 and 12 were reported previously.10,21
Sch em e 1. Syn th esis of Liga n d s 1-10a
a
(i) n-BuLi, TMEDA, 0 °C, (ii) chlorophosphine, -78 °C.
secondary amine functionality of phenoxazine was
protected using chloro-tert-butyldimethylsilane. Lithia-
tion of the protected backbone followed by the reaction
with the chlorophosphine and subsequent deprotection
gave the ligands in moderate yields after crystallization
(30%-43%). For the synthesis of 8 phenyllithium in-
stead of n-butyllithium was used as metallating agent
in order to prevent side product formation by nucleo-
philic substitution of the phenyl of the phosphine by
n-butyllithium.
F igu r e 3. Naphos-type ligands suitable for hydroformy-
lation of internal alkenes.
moieties were synthesized (1-10). The ligand backbones
were prepared according to literature procedures or
were commercially available. Selective dilithiation of the
backbones followed by the reaction with the chlorophos-
phine yields the corresponding ligands in moderate to
high yields (30-80%) (Scheme 1). Accidentally, in 31P-
{1H} NMR no distinction between the two different
phosphorus signals of 6 could be made, but they lose
their degeneracy in the (6)Rh(CO)2H complex.
Molecular mechanics calculations of ligands 1-10
show the effect of the heterocycle on the calculated
natural bite angle and the flexibility range (Table 1).5
The calculations show that the rigid 2,8-dimethylphen-
Ligands with the phenoxazine backbone (7 and 10)
were obtained via a slightly modified procedure. The

Hydroformylation of Internal Olefins
Organometallics, Vol. 22, No. 25, 2003 5361
Ta ble 1. Ca lcu la ted Na tu r a l Bite An gles a n d F lexibility Ra n ges for th e P h en oxa p h osp h in o-Mod ified
Xa n tp h os-Typ e Liga n d s
ligand
backbone
âna (deg)
flexibility rangea (deg)
1
2
3
4
5
6
7
8
10,11-dihydrodibenzo[b,f]oxepine
2,7-di-tert-butyl-10,10-dimethylphenoxasilin
2,7-dimethylphenoxathiin
10-isopropylidenexanthene
9,9-dimethylxanthene
105.7
108.0
112.5
114.2
116.0
124.3
124.5
131.2
111.8
128.9
123.1
121.4
91-128
106-137
107-140
110-144
100-134
117-156
113-151
126-153
98-133
benzo[k,l]xanthene
phenoxazine
10-phenylphenoxaphosphine
2,7-dimethylphenoxathiin
phenoxazine
2,7-di-tert-butyl-9,9-dimethylxanthene
2,7-di-tert-butyl-9,9-dimethylxanthene
9
10
11b
12b
117-145
107-142
107-138
a
The natural bite angle (ân) and the flexibility range were calculated as by Casey and Whiteker. ân is defined as the preferred chelation
angle determined only by backbone constraints and not by metal valence angles. The flexibility range is defined as the accessible range
of bite angles within 3 kcal mol-1 excess strain energy from the calculated natural bite angle. See also refs 10, 21.
b
F igu r e 4. Effect of phosphine moiety on bite angle and flexibility range (ân, flexibility range).
Ta ble 2. Selected High -P r essu r e NMR Da ta for
oxaphosphino moiety causes an increase of the bite
(Dip h osp h in e)Rh H(CO)2 Com p lexesa
angle compared to the corresponding diphenylphosphino
2
1J (Rh,H), 1J (Rh,P), J av(P,H), 2J (P,P),
δ
31P{1H} δ1H
(ppm) (ppm)
analogues, while the introduction of the methyl sub-
stituents on the phenoxaphosphino moiety causes a
decrease in bite angle compared to the unsubstituted
phenoxaphosphine (Figure 4). Ligand 8 has a calculated
natural bite angle that is much wider than expected
from the ligand backbone structure, as calculations
performed on the diphenylphosphino analogue show a
much smaller bite angle. In the case of 8 the phenyl
substituent in the backbone structure is in close prox-
imity of one of the methyl substituents on each phen-
oxaphosphino moiety, thereby forcing the phenoxaphos-
phino moieties to adopt a different orientation from that
initially expected, which results in an increased bite
angle.
ligand
Hz
Hz
Hz
Hz
1
2
3
4
11.9
124/129
150
153
154
151
153/149
153
153
42
-20.4
-23.7
-25.3
-27.2
-29.0
-33.0
-30.8
-25.7
-8.9
-8.7
-8.9
-9.0
-9.0
-9.2
-9.2
-8.9
n.o.b
6.6
12.4
12.5
7.7
6.6
13.5
bs
2.9
2.8
n.o.
n.o.
n.o.
n.o.
5
6c
7d
8
63
a
Conditions: p(CO/H2)(1:1) ) 20 bar, solvent ) toluene-d8, T
b
d
) 273 K. n.o.: not observed. c Solvent ) toluene-d8/THF. Sol-
vent ) THF-d8.
The complex (1)Rh(CO)2H was observed in lower quan-
tities, showing two broad doublets, indicating mainly
ae coordination. (9-10)Rh(CO)2H were not observed at
all. For (6)Rh(CO)2H a double doublet per phosphorus
atom was observed due to the inequivalence of the two
phosphorus atoms. In the 1H NMR at 273 K the hydride
appeared in the range from -9.2 to -8.7 ppm as a
triplet of doublets for ligands 1, 3, and 4, a triplet for 2
and 5-7, or a broad (distorted) singlet for 8. The
characteristic heteronuclear coupling constants and
chemical shifts are shown in Table 2.
Probably due to the high concentration that is re-
quired for high-pressure NMR, often small amounts of
additional rhodium-phosphine species were observed.
Both 1H NMR and 31P{1H} NMR using 13CO exclude
these species as either hydride- or carbonyl-containing
complexes. For ligand 8 many other phosphorus-
rhodium species, besides formation of (8)Rh(CO)2H, are
(Dip h osp h in e)Rh (CO)2H Com p lexes. To deter-
mine whether the variation in ligand backbone has an
influence on the chelation behavior of the ligands, we
studied the solution structures of the (diphosphine)Rh-
(CO)2H complexes, the resting state of the catalyst
under hydroformylation conditions, by high-pressure IR
and high-pressure NMR spectroscopy. The complexes
were prepared in situ from Rh(CO)2(acac) and diphos-
phine (1.1 equiv) under 16 bar of CO/H2 (1:1). For
ligands 1-8 the formation of (diphosphine)Rh(CO)2H
was observed to occur within 1 h at 40 °C. The 31P{1H}
NMR spectrum at 273 K of the complexes with 2-5, 7,
and 8 showed a characteristic doublet in the range from
-33 to -20 ppm. For ligands 1, 9, and 10 [(diphos-
phine)Rh(CO)2]2 were observed as the major complexes.
These dimers exhibit a splitting pattern in 31P{1H}
NMR consistent with an AA′A′′A′′′XX′ spin system at
- 27.4 ppm for 1, 4.8 ppm for 9, and -0.4 ppm for 10.36
(36) Castellanos-Paez, A.; Castillon, S.; Claver, C.; van Leeuwen,
P. W. N. M.; de Lange, W. G. J . Organometallics 1998, 17, 2543-2552.

5362 Organometallics, Vol. 22, No. 25, 2003
Bronger et al.
Ta ble 3. Selected High -P r essu r e IR Da ta for
(Dip h osp h in e)Rh H(CO)2 Com p lexesa
Ta ble 4. Hyd r ofor m yla tion of 1-Octen e a t 80 °Ca
ligand
Xb
ânb (deg) TOFc,d
l:bc
% isomc selc,e
b
b
b
b
ligand
ν1
ν2
ν3
ν4
1
2
3
4
5
6
7
8
C2H4
Si(CH3)2
S
isoprop
C(CH3)2
fused ring
NH
PPh
S
NH
C(CH3)2
C(CH3)2
105.7
108.0
112.5
114.2
116.0
124.3
124.5
131.2
111.8
128.9
123.1
121.4
1900
2250
3275
1800
1800
1600
1550
1100
343
5.5
14
22
23
23
30
33
18
2.8
6.8
10.5
9.7
10.0
12.5
11.6
7.5
14.2
11.5
10.0
10.8
82
87
85
86
86
84
85
87
85
87
89
88
1
2
2040
2041
2042
2037
2040
2036
2040
2038
1999
2000(vw)
1978
1981
1985
1979
1983
1979
1984
1980
1956
1953(vw)
3
4c
5
6c
7
8c
2004
1954
9
>99
>99
67
10
11
12
144
1557
453
a
Conditions: p(CO/H2)(1:1) ) 20 bar, solvent ) cyclohexane,
b
T ) 313 K. Vibrational bands belonging to either the ee or ae
62
configuration were assigned by performing a H-D exchange.
a
cSolvent ) 2-methyltetrahydrofuran.
Conditions: p(CO/H2)(1:1) ) 20 bar, ligand/Rh ) 5, substrate/
Rh ) 637, [Rh] ) 1.00 mM in toluene, number of experiments is
b
3. In none of the experiments was hydrogenation observed. X )
observed in considerable amounts in 31P{1H) NMR. The
relatively high concentration of ligand and rhodium that
is necessary for high-pressure NMR when compared to
the conditions of a catalytic reaction is a plausible cause
for the observation of so many species in which coordi-
nation of the additional phosphine of the ligand back-
bone can be involved. The disappearance of the signal
of the free phosphorus of the ligand backbone at -53
ppm indicates that this phosphorus atom is indeed
coordinating, reinforcing this hypothesis. The distorted
refers to the top bridging group in the ligand (see Figure 2); ân
)
the natural bite angle (see Table 1). c Linear over branched ratio,
percent linear aldehyde, percent isomerization to 2-octene, and
turnover frequency were determined at ∼ 20% alkene conversion.
Turnover frequency ) (mol aldehyde)‚(mol Rh)-1‚h-1 e Percent-
.
d
age of linear aldehyde of all products (including isomerization
toward internal alkenes).
sorption bands for the bridging carbonyl ligands (1801
cm-1/1733 cm-1), also absorptions at 2066 cm-1 (m) and
1994 cm-1 (s) were observed. In addition none of the
CO bands shifted upon using D2 instead of H2. Clearly
some other complexes are present when this ligand is
used. This complex was not stable since the carbonyl
absorptions diminished and many other absorptions
with a low intensity appeared after overnight measure-
ment at 80 °C.
Hyd r ofor m yla tion of 1-Octen e. Hydroformylation
of 1-octene was carried out at 80 °C under 20 bar of 1:1
CO/H2 using a 1.0 mM solution of rhodium diphosphine
catalyst prepared from Rh(CO)2(acac) and 5 equiv of
ligand. The production of octene isomers, nonanal, and
2-methyloctanal was monitored by gas chromatography.
Turnover frequencies were determined and averaged
after ∼20% conversion. The results of the experiments
are presented in Table 4.
In contrast to previous studies on the effect of bite
angle on selectivity for the hydroformylation of 1-octene,4
hardly any effect of the bite angle on the selectivity for
the linear aldehyde is observed for ligands 1-8. The l/b
ratio and the percentage of isomerization to internal
octenes increases, but both increase with increasing bite
angle up to a bite angle of 125° only. The observed
increase in l/b ratio is attributed to an increased
tendency of the branched alkyl rhodium species to
undergo â-hydrogen elimination instead of alkyl migra-
tion.32 After an initial increase in catalytic activity with
increasing bite angle from 105.7° up to 112.5° a drop in
activity with further increasing bite angle is observed.
Ligand 3, which according to the reported Hammet
constant37 has the strongest electron-withdrawing ca-
pacities with the sulfur moiety (σm(SPh) ) 0.2337) in the
ligand backbone, gives a catalytic system that shows
almost twice the activity compared to the results
obtained when the other phenoxaphosphino-modified
ligands are employed. These results show that small
structural and electronic differences can have a huge
impact on catalysis results.
1
broad signal at -8.9 ppm in H NMR might be the result
of two overlapping signals, which would indicate the
presence of more than one rhodium-hydride species.
Even at 183 K the signals could not be separated.
Formation of telomeric, oligomeric, or polymeric struc-
tures is not excluded. Characterization of the different
species has not been attempted.
The IR frequencies of the absorption bands of the (1-
8)Rh(CO)2H complexes in the carbonyl region are sum-
marized in Table 3. For the rhodium complexes with
ligands 3-7 only two absorption bands were observed.
By performing a hydride-deuterium exchange the two
bands shifted to a lower frequency. This indicates that
the ligands are coordinated in an ee fashion, as in this
isomer a trans relationship between the hydride and
carbonyl ligand exists, which results in coupling of the
vibrations. By exchange of H for D this resonance
interaction becomes much smaller, which leads to a shift
of the carbonyl bands of the ee isomer.3 Only when
ligands 1, 2, and 8 were used was the existence of an
equilibrium between ee and ae coordinating rhodium
complexes observed, as complexes with these ligands
showed all four absorption bands in the carbonyl
frequency region. Unlike the high-pressure NMR ex-
periments, no formation of the dimeric [(1)Rh(CO)2]2
was observed. This is ascribed to the low concentrations
that are used for high-pressure IR and which are similar
to those of catalytic runs, while high-pressure NMR
needs much higher concentrations, which favors the
formation of dimeric rhodium species.
For 10 only the dimeric complex [(10)Rh(CO)2]2 was
observed in the IR spectrum. This complex exhibits
absorption bands for the terminal carbonyl ligands
(2004 cm-1/1980 cm-1) and two absorption bands in the
bridging carbonyl ligands (1792 cm-1/1762 cm-1). Simi-
lar dimeric complexes were also observed for 12.10,21
[(12)Rh(CO)2]2 was reported to be the resting state of
the catalyst. For 9 clearly other species were also
formed. Besides two absorption bands for the terminal
carbonyl ligands (2010 cm-1/1985 cm-1) and two ab-
Very high regioselectivities and high percentages of
isomerization are observed for the dibenzophosphole
(37) Hansch, C.; Leo, A.; Taft, W. Chem. Rev. 1991, 91, 165-195.

Hydroformylation of Internal Olefins
Organometallics, Vol. 22, No. 25, 2003 5363
F igu r e 5. Turnover frequency (triangles) and regioselectivity (squares) vs the calculated natural bite angle in the
hydroformylation of trans-2-octene (ligands 1-8).
Ta ble 5. Hyd r ofor m yla tion of tr a n s-2-Octen e a t
120 °Ca
ligand
Xb
ânb (deg)
TOFc,d
l:bc
selc,e
1
2
3
4
5
6
7
8
C2H4
Si(CH3)2
S
isoprop
C(CH3)2
fused ring
NH
PPh
S
NH
C(CH3)2
C(CH3)2
105.7
108.0
112.5
114.2
116.0
124.3
124.5
131.2
111.8
128.9
123.1
121.4
250
193
151
143
143
63
0.4
29
2.6
4.4
5.8
4.3
8.0
8.3
3.7
72
81
85
81
89
89
79
38
8
9
78 (64f)
60f
18.3 (22.7f) 95 (96f)
10
11
12
24.7f
6.6 (9.2f)
9.5f
96f
140 (112f)
65f
87 (90f)
90f
a
Conditions: p(CO/H2)(1:1) ) 3.6 bar, ligand/Rh ) 10, substrate/
Rh ) 640, [Rh] ) 1.00 mM in toluene, number of experiments is
b
2. In none of the experiments was hydrogenation observed. X )
refers to the top bridging group in the ligand (see Figure 2); ân
)
the natural bite angle (see Table 1). c Linear over branched ratio,
percent linear aldehyde, and turnover frequency were determined
d
after 2 h. Turnover frequency ) (mol aldehyde)‚(mol Rh)-1‚h-1
.
e Percentage of linear aldehyde of all products other than octenes.
f p(CO/H2)(1:1) ) 2.0 bar.
F igu r e 6. Parity plot 1-octene vs 2-octene hydroformyla-
tion with ligands 1-8.
ligands 9 and 10. For both ligands the hydroformylation
activity is very low, but this can be a result of preferred
formation of the rhodium-dimer resting state.
Contrary to the results obtained with the noncyclic
analogues for the hydroformylation of 1-octene,4 an
increased bite angle leads to a strong decrease in
hydroformylation activity. While the activity decreases,
an increase in regioselectivity is observed that correlates
with the regioselectivity for the hydroformylation of
1-octene; see Figure 6 for a parity plot. As it is expected
that the rate of isomerization is crucial for obtaining
good catalysts for the selective hydroformylation of
internal alkenes to linear aldehydes, it is not surprising
that the ligands showing the highest rate of isomeriza-
tion in the hydroformylation of 1-octene show the
highest regioselectivity in the hydroformylation of trans-
2-octene. In addition, ligands with a higher rate of
isomerization show a lower hydroformylation activity,
as â-hydrogen elimination is nonproductive.
Hydr ofor m ylation of tr a n s-2-Octen e. Hydroformy-
lation of trans-2-octene was carried out at 120 °C under
3.6 bar of 1:1 CO/H2 using a 1.0 mM solution of rhodium
diphosphine catalyst prepared from Rh(PPh3)3H(CO)
and 10 equiv of ligand. The relatively high temperature
and low pressure are necessary to enhance the rate of
isomerization,38 thereby continuously maintaining the
formation of terminal alkenes. The production of alkene
isomers and linear and branched aldehydes was moni-
tored by gas chromatography. Turnover frequencies
were determined after 2 h reaction time.
The results of the hydroformylation experiments are
summarized in Table 5.
By comparing ligands 1-8 a clear bite angle effect
on catalytic activity and regioselectivity is observed in
the hydroformylation of trans-2-octene; see Figure 5.
Unprecedented high regioselectivities at moderate
activities for the hydroformylation of trans-2-octene are
obtained when the dibenzophoshole-type ligands 9 and
10 are used. These results can be attributed to the
combination of the high isomerization activity and the
(38) van Rooy, A.; Orij, E. N.; Kamer, P. C. J .; van Leeuwen, P. W.
N. M. Organometallics 1995, 14, 34-43.

5364 Organometallics, Vol. 22, No. 25, 2003
Ta ble 6. Deu ter iofor m yla tion of 1-Hexen ea
1-hexene (%) 2-hexene-d1 (%)
Bronger et al.
conversion to
2-methyl-
ligand aldehydes (%) CH2dCHC4H9 CH2dCDC4H9 CHDdCHC4H9 CHD-CHdCHC3H7 heptanal-d2 (%)b hexanal-d2 (%)b
1
5
7
20.7
23.8
13.9
77.1
71.5
83.2
1.34
0.56
0.68
0.29
1.76
1.13
0.6
2.4
1.6
17.2
22.7
13.4
3.5
1.1
0.5
a
Conditions: p(CO/D2) ) 20 bar (∼ 1:1, see Experimental Section for more details), ligand/Rh ) 5, substrate/Rh ) 806, [Rh] ) 1.00
mM in toluene. In none of the experiments was hydrogenation observed. It is important to note that the amounts of deuterated hexenes
do not reflect the actual percentages of formed (1- and 2-alkyl)rhodium intermediates, as also nonproductive â-D elimination occurs.
b
Deuterium labels were almost exclusively found in the formyl group and the position â to the formyl group. Only traces of deuterium
label was found at the position R to the formyl group.
Ta ble 7. P a r tition in g of (Alk yl)r h od iu m In ter m ed ia tes
(1-alkyl) intermediatea
(2-alkyl) intermediateb
(1-alkyl/2-alkyl)
intermediate
ligand
l/b
heptanal (%) CH2dCDC4H9 (%) 2-methylhexanal (%) CDHdC5H10 (%) CDH2CHdC4H8 (%)
1
5
7
4.22
4.42
4.37
4.9
20
25
92.8
97.6
95.2
7.2
2.4
4.8
79.7
20.9
15.5
6.6
33.5
35.0
13.7
45.6
49.5
a
b
CH2dCDC4H9 + heptanal ) 100% (data converted from Table 6). CDHdC5H10 + CDH2CHdC4H8 + 2-methylhexanal ) 100% (data
converted from Table 6).
very high regioselectivity of these ligands. Compared
to ligands 1-8 only a small difference in hydroformy-
lation activity between the 1-octene and trans-2-octene
is observed, especially for 10.
directly. On the other hand, the branched alkyl inter-
mediate partitions between conversion toward the
branched aldehyde, 1-hexene, and cis- and trans-2-
hexene.
In an attempt to improve the regioselectivity even
further, the effect of ligand concentration on activity and
selectivity was investigated for 9. Upon increasing the
ligand:rhodium ratio from 5 to 40 equiv, a gradual
decrease in TOF from 75 mol‚mol-1‚h-1 to 34 mol‚
mol-1‚h-1 was observed. Within this range the regiose-
lectivity did not change. This indicates a very strong
coordination of the diphosphine to the rhodium metal;
the formation of an inactive (diphosphine)2Rh species
is favored at increased ligand:rhodium ratios.
From these results it can be concluded that increasing
the bite angle in a branched alkyl rhodium intermediate
causes a decrease in branched aldehyde formation, while
the percentage of isomerization toward both 1-hexene
and 2-hexene increases.
Discu ssion
Sp ectr oscop y. The high-pressure NMR data of the
(1-8)RhH(CO)2 and [(9-10)Rh(CO)2]2 complexes are in
line with the results obtained with high-pressure IR.
For (1-8)RhH(CO)2 the small rhodium-proton coupling
constant and phosphorus-proton coupling constant and
the large rhodium-phosphorus coupling constant indi-
cate that the dynamic equilibrium between ee and ae
isomers is shifted toward the ee isomer. Quantitative
estimations of the ee:ae ratios for (1-2, 8)RhH(CO)2
cannot be made using the averaged phosphorus-proton
coupling constants, as the ae and ee phosphorus-proton
coupling constants are not known for these phosphacy-
clic diphosphines. As is indicated in Table 2, they can
differ substantially between the different ligands; for
example, a relatively large phosphorus-proton coupling
is observed for complexes with ligands 3 and 7 and a
small phosphorus-proton coupling is observed for the
complex with ligand 2. High-pressure IR indicates
exclusive formation of ee isomers for the former two
complexes and an ee-ae mixture for the latter complex.
In nondisturbed trigonal bipyramidal complexes the
phosphorus-hydride coupling should be very small;4
therefore the results imply that in most cases the
trigonal bipyramidal rhodium complexes are highly
distorted, with the phosphines bent out of the equatorial
plane. The rhodium-proton coupling constants are also
indicative of the ee:ae ratio. The ae-coordinated isomers
often have large rhodium-proton coupling constants,
while for purely ee-coordinating ligands, like BISBI,
small coupling constants are observed.39 Moreover, in
previous studies increasing ee:ae ratios were accompa-
nied by decreasing rhodium-proton coupling con-
Deu ter iofor m yla tion of 1-Hexen e. Deuterioformy-
lation of 1-hexene was carried out under conditions
similar to the hydroformylation experiments. 1-Hexene
was used for the deuterioformylation instead of 1-octene
1
to facilitate comparison with literature data of the H
2
and H spectra.32 The percentages of deuterated species
formed in the deuterioformylation were calculated using
2
GC and H{1H} NMR data on the ratio of heptanal to
branched aldehydes, percent conversion, percent isomer-
ization toward internal hexenes, and deuterium con-
tents of aldehydes and recovered hexenes. Activities
were comparable to the 1-octene hydroformylation
experiments, while the regioselectivities are somewhat
lower than those obtained with 1-octene. Table 6 sum-
marizes the deuterioformylation results.
Previous studies by Casey,32 Lazzaroni,33,34 Pino,35
and van Leeuwen28 have shown that the formation of
the alkylrhodium intermediate is reversible at elevated
temperatures. By a similar experimental method, the
amount of initial linear alkyl rhodium formation and
branched alkyl rhodium formation can be easily de-
duced. Detailed dissection of the labeling results of
catalysis with 1, 5, and 7 shows that the initial ratio of
linear-alkyl to branched-alkyl intermediates is ∼4.2-
4.4 (Table 7). These results prove that when these
ligands are used, the differences in observed l/b ratio
are not only the result of preferential formation of the
linear aldehyde, but mainly the result of different rates
of â-hydrogen elimination. In all three cases >92% of
the linear alkyl intermediate reverts to the aldehyde

Hydroformylation of Internal Olefins
Organometallics, Vol. 22, No. 25, 2003 5365
Sch em e 2. â-Hyd r ogen Elim in a tion vs CO
Associa tion
F igu r e 7. Effect of bite angle on space around rhodium.
stants.3,4 It is surprising that for (2)RhH(CO)2 the
rhodium-proton coupling could not be resolved (1J (Rh,H)
< 2 Hz) even at low temperatures (183 K), while high-
pressure IR clearly shows absorption bands belonging
to the ae isomer, albeit in a small amount.
to undergo â-hydrogen elimination (B f A) than CO
association (B f C), as required for hydroformylation.
From the deuterioformylation experiments it can be
concluded that isomerization toward internal olefins is
faster than isomerization toward the terminal olefin.
Also, the rate of isomerization is faster than the rate of
CO insertion. Additionally, CO insertion into the
branched alkyl rhodium intermediate is slower than CO
insertion into the linear alkyl rhodium intermediate;
especially when bulky ligands are used, this CO inser-
tion is hampered. Thus, it is reasoned that systems that
show a higher rate of isomerization (thus ligands with
wide bite angles) show a lower hydroformylation activ-
ity. For 1-octene hydroformylation these effects are less
pronounced, as differences between the rate of â-hydro-
gen elimination from the linear alkyl rhodium species
are smaller and since CO association to a linear alkyl
rhodium intermediate is sterically less demanding.
Another explanation for the bite angle effect on the
activity in hydroformylation is found in a decreased rate
of alkene insertion. The increased steric congestion
around rhodium hampers the olefin from entering the
coordination sphere. For internal olefins this effect is
stronger than for terminal olefins.
Ca t a lysis: R egioselect ivit y. Concerning regio-
selectivity, the series does follow the trend that was
observed before;4 that is, ligands possessing larger bite
angles lead to more regioselective systems. It must be
noted that the origin of the regioselectivity is different
from that reported previously in which the increase in
selectivity is attributed to more selective formation of
the linear alkyl-rhodium species. In the present study
using 1-8 the differences in regioselectivity are caused
by the differences in rate of â-hydrogen elimination of
the branched alkyl intermediate compared to the linear
alkyl intermediate, as supported by the deuterioformy-
lation experiments. The branched alkyl intermediate
reverts to an olefin complex, but the linear alkyl
intermediate is converted directly to the aldehyde.32 As
more linear than branched products are formed for
hydroformylation of trans-2-octene using 2-8, it can be
concluded that most products are formed via in situ
generated terminal olefins. The order in the regioselec-
tivities of the different systems is therefore unchanged.
It can be concluded that the formation of terminal
olefins from internal olefins is faster than the formation
of branched aldehydes.
The results obtained by high-pressure IR suggest that
the electron density around rhodium is similar for all
(1-8)RhH(CO)2 complexes, as all bands are within close
range. Comparison of the CO-stretch frequencies with
the (diphosphine)RhH(CO)2 complexes of the diphe-
nylphosphine analogues of ligands 1-84 shows that all
ligands have similar electronic properties. The large
differences in catalytic activity between these two
systems are therefore attributed to extended π-conjuga-
tion21 and increased effective steric bulk upon introduc-
tion of the phenoxaphosphino moiety and are to a lesser
extent the result of decreased phosphorus basicity.
Ca ta lysis: Activity. Widening the bite angle in the
square-planar (diphosphine)Rh(CO)H complex, one of
the supposed intermediates in the hydroformylation
reaction would certainly accelerate the overall reaction
by increasing the concentration of this intermediate
complex, as a trans coordination of the diphosphine is
energetically more favorable than a cis coordination. If
other effects are absent or counterbalance one another
(i.e., alkene and CO association/dissociation, hydride
and alkyl migration, and hydrogenolysis), then this
leads to an increase in hydroformylation activity, as was
postulated previously.4 In the same study an increased
rate of alkene coordination for ligands with a wider bite
angle was also proposed in order to explain the effect
on reaction rate in accord with other studies.40,41 While
these positive effects of widening the bite angle on
activity are certainly not ruled out for 1-8, the results
indicate that another step in the catalytic cycle becomes
rate-limiting. This effect is especially pronounced in the
hydroformylation of trans-2-octene. The decrease of
hydroformylation activity with increased bite angle
might be explained by a step later in the catalytic cycle
becoming rate limiting. In the trigonal bipyramidal
rhodium complex, increasing the bite angle results in
increased steric congestion around rhodium especially
in the apical position. In addition, with the exception
of 2 and 7 an increased bite angle leads to a decreased
phosphorus-hydride coupling, which is indicative of
approaching a more perfect trigonal bipyramidal struc-
ture. This might be an indication that ligands with a
larger bite angle lead to less distorted trigonal bipyra-
midal complexes and thereby to an increased effective
steric bulk (Figure 7).
Upon increasing the natural bite angle to 131° with
ligand 8, a decrease in regioselectivity is observed.
Although this might be due to possible coordination of
the phosphine in the ligand backbone, thereby acting
partially as a monophosphine, it is unlikely that this is
happening, since at low phosphorus concentrations and
under CO pressure the monophosphine would dissoci-
ate. A more plausible reason is that systems possessing
too large bite angles again lead to less regioselective
systems by relatively increasing the rate of â-hydrogen
According to this postulation, it is more favorable for
complexes of B (Scheme 2) with wide bite angle ligands
(39) Bregman, F. R.; Ernsting, J . M.; Mu¨ller, F.; Boele, M. D. K.;
van der Veen, L. A.; Elsevier, C. J . J . Organomet. Chem. 1999, 592,
306-311.
(40) Rocha, W. R.; De Almeida, W. B. Organometallics 1998, 17,
1961-1967.
(41) Thorn, D. L.; Hoffmann, R. J . Am. Chem. Soc. 1978, 100, 2079-
2090.

5366 Organometallics, Vol. 22, No. 25, 2003
Bronger et al.
thene,43,44 10,11-dihydrodebenzo[b,f]oxepine,45 9-isopropyli-
denexanthene,46 10-phenylphenoxaphosphine,47 2,8-dimethyl-
10-chlorophenoxaphosphine,48 2,7-di-tert-butyl-10,10-dimeth-
ylphenoxasilin,49,50 2,7-dimethylphenoxathiin,51 9-(tert-but-
yldimethylsilyl)phenoxazine,52 2,7-di-tert-butyl-9,9-dimethyl-
4,5-bis(10-phenoxaphosphino)xanthene (11),10,21 and 4,5-bis-
(9-dibenzo[b,d]phospholyl)-2,7-di-tert-butyl-9,9-dimethylxan-
thene (12)10,21 were prepared according to literature procedures.
Silica gel 60 (230-400 mesh) purchased from Merck was used
for column chromatography. Melting points were determined
on a Gallenkamp MFB-595 melting point apparatus in open
capillaries and are reported uncorrected. NMR spectra were
recorded on a Varian Mercury 300 or Inova 500 spectrometer.
31P and 13C spectra were measured 1H decoupled. TMS was
used as a standard for 1H and 13C NMR and 85% of H3PO4 in
H2O for 31P NMR. Infrared spectra were recorded on a Nicolet
510 FT-IR spectrophotometer. High-pressure IR spectra were
measured using a 50 mL homemade stainless steel autoclave
equipped with mechanical stirrer and ZnS windows. Hydro-
formylation reactions were carried out in a 200 mL homemade
stainless steel autoclave. Both high-pressure IR and hydro-
formylation experiments were conducted at a stirring rate of
800 rpm. The alkene was filtered over neutral activated
alumina to remove peroxide impurities. The reactions were
stopped by quenching the reaction mixture with tri-n-butyl
phosphite, cooling on ice, and venting the gases. Synthesis gas
(CO/H2, 1:1, 99.9%) was purchased from Air Liquide. D2 was
purchased from Hoekloos. 13CO (99%) was purchased from
Cambridge Isotope Laboratories. Gas chromatographic analy-
ses were run on an Interscience HR GC Mega 2 apparatus
(split/splitless injector, J &W Scientific, DB-1 30 m column, film
thickness 3.0 mm, carrier gas 70 kPa He, FID detector)
equipped with a Hewlett-Packard data system (Chrom-Card)
using decane as an internal standard.
elimination from the linear alkyl intermediate. An
optimum bite angle is found around 120-125°. Previous
studies on a dibenzofuran-based diphenylphosphine
(4,6-bis(diphenylphosphinodibenzofuran, ân ) 131.1°)
also showed a decrease in selectivity compared to
similar ligands with a smaller bite angle.11
The results with the dibenzophosphole-type ligands
9, 10, and 12 show that very subtle changes in either
backbone structure or cyclic phosphino moiety can have
a tremendous influence on the catalysis results. How
these subtle changes influence catalysis is not yet
understood. Unfortunately, the preferential formation
of the rhodium dimer complex hampers a detailed in
situ study.
Con clu sion s. Within the series of xantphenoxaphos-
type ligands no distinct effect of the natural bite angle
on the coordination mode of the (1-8)Rh(CO)2H com-
plexes has been found. Most ligands show ee coordina-
tion preferentially, and only for ligands 1, 2, and 8 was
also ae coordination observed. The decrease of catalytic
activity with an increase in natural bite angle is
attributed to differences in the rate of â-hydrogen
elimination and CO association in the intermediate
(1-8)Rh(CO)(alkyl) complex, and these are mainly
determined by steric factors. The differences in selectiv-
ity are also attributed to the differences in rate of
â-hydrogen elimination, as deuterioformylation experi-
ments have shown that the initial ratio of (linear alkyl)-
rhodium intermediate:(branched alkyl)rhodium inter-
mediate is equal when 1, 5, and 7 are used. To the best
of our knowledge, thixantphenoxaphos (3) gives an
unprecedented high activity with a respectable selectiv-
ity in the rhodium-diphosphine-catalyzed hydroformy-
lation of terminal olefins. Noteworthy, though, are the
recent results obtained by Stanley et al., who improved
existing systems by the addition of water to the reaction
medium.42
4,5-Bis(2,8-d im et h yl-10-p h en oxa p h osp h in o)-10,11-d i-
h yd r od iben zo[b,f]oxep in e (Hom oxa n tp h en oxa p h os, 1).
At 0 °C 1.9 mL of n-butyllithium (2.5 M in hexanes, 4.8 mmol)
was added to a stirred solution of 380 mg of 10,11-dihydrod-
ibenzo[b,f]oxepine (1.9 mmol) and 0.73 mL of TMEDA (4.8
mmol) in 10 mL of diethyl ether. The resulting solution was
warmed to room temperature overnight. The reaction mixture
was cooled to -78 °C, and a solution of 1.25 g of 10-chloro-
2,8-dimethylphenoxaphosphine (4.8 mmol) in 10 mL of toluene
was added. The reaction mixture was slowly warmed to room
temperature overnight. Next the diethyl ether was removed
in vacuo, and the mixture was diluted with 30 mL of dichlo-
romethane and hydrolyzed with 10 mL of a 10% aqueous HCl
solution. The water layer was removed, and the organic layer
was dried over MgSO4. The solvents were removed in vacuo,
and the resulting white solid was crystallized from 2-propanol/
toluene. Yield: 860 mg of white crystals (68%). Mp: 247-249
°C. 31P{1H} NMR (CDCl3): δ -69.12. 1H NMR (CDCl3): δ 7.67
(d, 3J (P,H) ) 1.5 Hz, 4H), 7.12 (dd, 3J (H,H) ) 8.0 Hz, 4J (H,H)
) 2.0 Hz, 4H), 7.07 (d, 3J (H,H) ) 8.5 Hz, 4H), 7.01 (dd, 3J (H,H)
High regioselectivities are observed when the diben-
zophosphole-type ligands 9 and 10 are used for the
hydroformylation of trans-2-octene, which are unprec-
edented. These regioselectivities even supersede the
regioselectivities often reported for the hydroformylation
of terminal olefins. Despite their moderate 1-octene
hydroformylation activity, 9 and 10 still show an ac-
ceptable activity in hydroformylation of trans-2-octene.
Exp er im en ta l Section
Com p u ta tion a l Deta ils. The molecular mechanics calcula-
tions were performed using CAChe WorkSystem version 4.0,
on an Apple Power MacIntosh 950, equipped with two CAChe
CXP coprocessors. Calculations were carried out similarly to
the method described by Casey and Whiteker,5 using a Rh-P
bond length of 2.315 Å. Minimizations were done using the
block-diagonal Newton-Raphson method, allowing the struc-
tures to converge with a termination criterion of a rms factor
of 0.0001 kcal mol-1 Å-1 or less.
Gen er a l P r oced u r e. All air- or water-sensitive reactions
were performed using standard Schlenk techniques under an
atmosphere of purified argon. Toluene was distilled from
sodium, THF from sodium/benzophenone, and hexanes from
sodium/benzophenone/triglyme. 2-Propanol and dichloromethane
were distilled from CaH2. Chemicals were purchased from
Acros Chimica and Aldrich Chemical Co. Benzo[k,l]xan-
4
3
) 7.0 Hz, J (H,H) ) 2.5 Hz, 2H), 6.89 (dd, J (H,H) ) 7.0 Hz,
4J (H,H) ) 1.5 Hz, 2H), 6.86 (t, 3J (H,H) ) 7.0 Hz), 3.07 (s, 4H),
(43) Orchin, M. J . Am. Chem. Soc. 1948, 70, 495-497.
(44) Yagupsky, G.; Wilkinson, G. J . Chem. Soc. (A) 1969, 725-733.
(45) Hess, B. A., J r.; Bailey, A. S.; Bartusek, B.; Boekelheide, V. J .
Am. Chem. Soc. 1969, 91, 1665-1672.
(46) McMurry, J . E.; Krepski, L. R. J . Org. Chem. 1976, 41, 3929-
3930.
(47) Schlosser, M. In Organometallics in Synthesis: A Manual;
Schlosser, M., Ed.; Wiley: Chichester, 1994; p 115.
(48) Freedman, L. D.; Doak, G. O.; Edmisten, J . R. J . Org. Chem.
1961, 26, 284-285.
(49) Gilman, H.; Trepka, W. J . J . Org. Chem. 1961, 26, 5202.
(50) Oita, K.; Gilman, H. J . Am. Chem. Soc. 1957, 79, 339-342.
(51) Suter, C. M.; McKenzie, J . P.; Maxwell, C. E. J . Am. Chem.
Soc. 1936, 58, 717-720.
(52) Antonia, Y.; Barrera, P.; Contreras, O.; Franco, F.; Galeazzi,
E.; Garcia, J .; Greenhouse, R.; Guzman, A.; Velarde, E.; Muchowski,
J . M. J . Org. Chem. 1948, 54, 2159-2165.
(42) Aubrey, D. A.; Bridges, N. N.; Ezell, K.; Stanley, G. G. J . Am.
Chem. Soc. 2003, 125, 11180-11181.

Hydroformylation of Internal Olefins
Organometallics, Vol. 22, No. 25, 2003 5367
2.24 (s, 12H). 13C{1H} NMR (CDCl3): δ 158.01 (t, 10.9 Hz),
153.23 (s), 135.10 (t, 18.7 Hz), 132.86 (t, 16.6 Hz), 132.68 (s),
132.449 (t, 4.9 Hz), 131.34 (s), 131.02 (s), 123.91 (s), 118.83
(s), 117.24 (s), 32.58 (s), 20.51 (s). Anal. Calcd for C42H34O3P2:
C, 77.77; H, 5.28. Found: C, 77.95; H, 5.40.
was dried over MgSO4. The solvents were removed in vacuo,
and the resulting yellow solid was precipitated from 2-pro-
panol/toluene. Yield: 1.5 g of a slightly tanned powder (38.8%).
1
Mp: 333-335 °C (dec). 31P{1H} NMR (CDCl3): δ -71.10. H
NMR (CDCl3): δ 7.86 (bd, 3J (P,H) ) 8.7 Hz, 4H), 7.22 (dd,
4
3
3J (H,H) ) 7.2 Hz, J (H,H) ) 2.7 Hz, 2H), 7.17 (dd, J (H,H) )
8.4 Hz, 4J (H,H) ) 2.1 Hz, 4H), 7.11 (d, 3J (H,H) ) 8.1 Hz, 4H),
6.92 (t, 3J (H,H) ) 7.5 Hz, 2H), 6.75 (bdd, 3J (H,H) ) 7.8 Hz,
4J (H,H) ) 1.8 Hz, 2H), 2.33 (s, 12H), 2.00 (s, 6H). Anal. Calcd
for C44H36O3P2: C, 78.33; H, 5.38. Found: C, 78.45; H, 5.53.
4,5-Bis(2,8-d im eth yl-10-p h en oxa p h osp h in o)-9,9-d im e-
th ylxa n th en e (Xa n tp h en oxa p h os, 5). At 0 °C 6 mL of
n-butyllithium (2.5 M in hexanes, 15 mmol) was added to a
stirred solution of 1 g of 9,9-dimethylxanthene (4.8 mmol) and
2.2 mL of TMEDA (15 mmol) in 20 mL of diethyl ether. The
resulting solution was warmed to room temperature and
stirred overnight. The reaction mixture was cooled to -78 °C,
and a solution of 3.9 g of 10-chloro-2,8-dimethylphenoxaphos-
phine (15 mmol) in 20 mL of toluene was added. The reaction
mixture was slowly warmed to room temperature and stirred
overnight. Next the diethyl ether was removed in vacuo, and
the mixture was diluted with 40 mL of dichloromethane and
hydrolyzed with 10 mL of a 10% aqueous HCl solution. The
water layer was removed, and the organic layer was dried over
MgSO4. The solvents were removed in vacuo, and the resulting
yellow-white solid was crystallized from 2-propanol/toluene.
Yield: 1.9 g of white crystals (59%). Mp: 328-329 °C. 31P-
{1H} NMR (CDCl3): δ -70.96. 1H NMR (CDCl3): δ 7.98 (d,
4,5-Bis(2,8-d im eth yl-10-p h en oxa p h osp h in o)-2,7-d i-ter t-
bu tyl-10,10-d im eth ylp h en oxa silin (Sixa n tp h en oxa p h os,
2). At 0 °C 3.6 mL of n-butyllithium (2.5 M in hexanes, 9.0
mmol) was added to a stirred solution of 1.23 g of 2,7-di-tert-
butyl-10,10-dimethylphenoxasilin (3.6 mmol) and 1.4 mL of
TMEDA (9.3 mmol) in 20 mL of diethyl ether. The resulting
solution was warmed to room temperature overnight. The
reaction mixture was cooled to -78 °C, and a solution of 2.4 g
of 10-chloro-2,8-dimethylphenoxaphosphine (9.1 mmol) in 10
mL of toluene was added. The reaction mixture was slowly
warmed to room temperature overnight. Next the diethyl ether
was removed in vacuo, and the mixture was diluted with 60
mL of dichloromethane and hydrolyzed with 20 mL of a 10%
aqueous HCl solution. The water layer was removed, and the
organic layer was dried over MgSO4. The solvents were
removed in vacuo, and the resulting yellow-white foam was
crystallized from 2-propanol/toluene. Yield: 1.3 g of white
crystals (44.4%). Mp: 300-302 °C. 31P{1H} NMR (CDCl3): δ
1
4
-72.09. H NMR (CDCl3): δ 8.06 (s, 4H), 7.42 (d, J (H,H) )
2.5 Hz, 2H), 7.22 (dd, 3J (H,H) ) 8.5 Hz, 4J (H,H) ) 2.0 Hz,
4H), 7.17 (d, 3J (H,H) ) 8.0 Hz, 4H), 7.17 (d, 4J (H,H) ) 2.0 Hz,
2H), 2.39 (s, 12H), 1.20 (s, 18H), 0.49 (s, 6H). 13C{1H} NMR
(CDCl3): δ 156.41 (t, 11.0 Hz), 151.13 (s), 142.05 (s), 132.57
(m), 131.35 (s), 129.54 (s), 128.41 (s), 128.28 (s), 126.12 (s),
125.42 (vt), 115.70 (s), 114.37 (s), 31.28 (s), 28.24 (s), 17.72
(s), -3.0 (s). Anal. Calcd for C50H52O3P2Si: C, 75.92; H, 6.63.
Found: C, 75.75; H, 6.65.
3
4
3J (P,H) ) 6.0 Hz, 4 H), 7.29 (dd, J (H,H) ) 5.0 Hz, J (H,H) )
1.5 Hz, 4H), 7.18 (dd, 3J (H,H) ) 8.5 Hz, 4J (H,H) ) 2.5 Hz,
3
3
4H), 7.11 (d, J (H,H) ) 8 Hz, 2H), 6.90 (t, J (H,H) ) 8.0 Hz,
2H), 6.75 (bd, 3J (H,H) ) 7.5 Hz, 2H), 2.35 (s, 12H), 1.55 (s,
6H). 13C{1H} NMR (CDCl3): δ 154.35 (s), 152.41 (t, 10.56 Hz),
135.77 (t, 21.5 Hz), 132.96 (t, 5.4 Hz), 132.09 (s), 131.74 (s),
130.26 (s), 127.79 (vt, 11.44 Hz), 127.18 (s), 123.76 (s), 118.16
(s), 117.66 (s), 34.58 (s), 32.59 (s), 20.85 (s). Anal. Calcd for
4,5-Bis(2,8-d im eth yl-10-p h en oxa p h osp h in o)-2,7-d im e-
th ylp h en oxa th iin (Th ixa n tp h en oxa p h os, 3). At 0 °C 11.0
mL of n-butyllithium (2.5 M in hexanes, 27.6 mmol) was added
to a stirred solution of 2.5 g of 2,7-dimethylphenoxathiin (11.0
mmol) and 4.2 mL of TMEDA (27.4 mmol) in 50 mL of diethyl
ether. The resulting solution was warmed to room temperature
overnight. The reaction mixture was cooled to -78 °C, and a
solution of 7.2 g of 10-chloro-2,8-dimethylphenoxaphosphine
(27.4 mmol) in 30 mL of toluene was added. The reaction
mixture was slowly warmed to room temperature overnight.
Next the diethyl ether was removed in vacuo, and the mixture
was diluted with 120 mL of dichloromethane and hydrolyzed
with 40 mL of a 10% aqueous HCl solution. The water layer
was removed, and the organic layer was dried over MgSO4.
The solvents were removed in vacuo, and the resulting yellow
solid was crystallized from 2-propanol/toluene. Yield: 3.3 g of
white crystals (43.5%). Mp: 333-338 °C (dec). 31P{1H} NMR
(CDCl3): δ -71.28. 1H NMR (CDCl3): δ 7.77 (bs, 4 H), 7.17
(dd, 3J (H,H) ) 8.4 Hz, 4J (H,H) ) 2.1 Hz, 4H), 7.08 (d, 3J (H,H)
) 8.4 Hz, 4H), 6.78 (bs, 2H), 6.46 (bs, 2H), 2.32 (s, 12H), 2.03
(s, 6H). 13C{1H} NMR (CDCl3): δ 153.89 (s), 152.61 (t, 12.2
Hz), 149.86 (s), 135.53 (t, 20.8 Hz), 134.61 (s), 132.93 (t, 4.8
Hz), 131.76 (s), 129.74 (t, 15.8 Hz), 128.37 (s), 120.42 (s), 117.97
(t, 3.6 Hz), 117.67 (s), 20.93 (s). Anal. Calcd for C42H34O3P2S:
C, 74.10; H, 5.03. Found: C, 73.48; H, 4.98.
4,5-Bis(2,8-d im e t h yl-10-p h e n oxa p h osp h in o)-9-iso-
p r op ylid en exa n th en e (Isop r op xa n tp h en oxa p h os, 4). At
0 °C 5.8 mL of n-butyllithium (2.5 M in hexanes, 14.5 mmol)
was added to a stirred solution of 1.3 g of 10-isopropyliden-
exanthene (6.8 mmol) and 2.25 mL of TMEDA (14.6 mmol) in
30 mL of diethyl ether. The resulting solution was warmed to
room temperature overnight. The reaction mixture was cooled
to -78 °C, and a solution of 3.8 g of 10-chloro-2,8-dimeth-
ylphenoxaphosphine (14.5 mmol) in 10 mL of toluene was
added. The reaction mixture was slowly warmed to room
temperature overnight. Next the diethyl ether was removed
in vacuo, and the mixture was diluted with 60 mL of dichlo-
romethane and hydrolyzed with 20 mL of a 10% aqueous HCl
solution. The water layer was removed, and the organic layer
C
43H36O3P2: C, 77.93; H, 5.48. Found: C, 77.91; H, 5.54.
6,7-Bis(2,8-d im eth yl-10-p h en oxa p h osp h in o)ben zo[k,l]-
xa n th en e (Ben zoxa n tp h en oxa p h os, 6). At 0 °C 0.8 mL of
n-butyllithium (2.5 M in hexanes, 2.0 mmol) was added to a
stirred solution of 180 mg of benzo[k,l]xanthene (0.82 mmol)
and 0.31 mL of TMEDA (2 mmol) in 10 mL of diethyl ether.
The resulting solution was warmed to room temperature
overnight. The reaction mixture was cooled to -78 °C, and a
solution of 525 mg of 10-chloro-2,8-dimethylphenoxaphosphine
(2.0 mmol) in 5 mL of toluene was added. The reaction mixture
was slowly warmed to room temperature overnight. Next the
diethyl ether was removed in vacuo, and the mixture was
diluted with 40 mL of dichloromethane and hydrolyzed with
10 mL of a 10% aqueous HCl solution. The water layer was
removed, and the organic layer was dried over MgSO4. The
solvents were removed in vacuo, and the resulting yellow solid
was crystallized from 2-propanol/toluene. Yield: 280 mg of
yellow crystals (51%). Mp: 308-310 °C. 31P{1H} NMR
(CDCl3): δ -70.86. 1H NMR (CDCl3): δ 8.14 (bd, 3J (H,H) )
6.5 Hz, 2H), 7.94 (bd, 3J (H,H) ) 6.5 Hz, 2H), 7.86 (dd, 3J (H,H)
) 8.0 Hz, 4J (H,H) ) 1.5 Hz, 1H), 7.70 (d, 3J (H,H) ) 7.0 Hz,
1H), 7.50 (d, 3J (H,H) ) 7.5 Hz, 1H), 7.40 (t, 3J (H,H) ) 7.5 Hz,
1H), 7.26-7.2 (m, 5H), 7.17-7.13 (m, 8.0 Hz, 4H), 6.95 (t, 8.0
3
Hz, 1H), 6.88 (dd, J (H,H) ) 8.5 Hz, 3J (H,H) ) 1.5 Hz, 1H),
6.73 (bdd, 3J (H,H) ) 7.5 Hz, 3J (H,H) ) 1.5 Hz, 1H), 2.39 (s,
6H), 2.37 (s, 6H). 13C{1H} NMR (CDCl3): δ 155.59 (s), 154.98
(s), 154.41 (t, 10.9 Hz), 136.58 (vt), 136.56 (s), 136.05 (vt),
134.81 (s), 134.06 (m), 132.87 (s), 132.55 (s), 131.76 (s), 129.56
(s), 129.18 (s), 128.25 (s), 126.65 (s), 125.03 (s), 124.81 (s),
122.17 (s), 121.72 (s), 121.63 (s), 119.40 (s), 119.15 (s), 118.96
(t, 2.7 Hz), 118.48 (s), 118.46 (s), 118.43 (s), 118.34 (s), 116.28
(s), 20.83 (s), 20.80 (s). Anal. Calcd for C44H32O3P2: C, 78.80;
H, 4.81. Found: C, 79.52; H, 5.01.
4,5-Bis(2,8-d im et h yl-10-p h en oxa p h osp h in o)p h en ox-
a zin e (Nixa n tp h en oxa p h os, 7). At 0 °C 13.8 mL of n-

5368 Organometallics, Vol. 22, No. 25, 2003
Bronger et al.
butyllithium (2.5 M in hexanes, 34.5 mmol) was added to a
stirred solution of 4.0 g of 9-(tert-butyldimethylsilyl)phenox-
azine (13.5 mmol) and 5.2 mL of TMEDA (34.5 mmol) in 200
mL of diethyl ether. The resulting solution was warmed to
room temperature overnight. The reaction mixture was cooled
to -78 °C, and a solution of 9.1 g of 10-chloro-2,8-dimeth-
ylphenoxaphosphine (34.5 mmol) in 60 mL of toluene was
added. The reaction mixture was slowly warmed to room
temperature overnight. Next the diethyl ether was removed
in vacuo, and the mixture was diluted with 120 mL of
dichloromethane and hydrolyzed with 40 mL of a 10% aqueous
HCl solution. The water layer was removed, and the organic
layer was dried over MgSO4. The solvents were removed in
vacuo, and the resulting brown paste was dissolved in 100 mL
of THF. n-(Bu)4NF‚3H2O (8.62 g, 33 mmol) was added, and
the reaction mixture was stirred for 2 days at room temper-
ature. Then 50 mL of brine and 200 mL of dichloromethane
were added. The water layer was removed, and the organic
layer was dried on MgSO4. The solvents were removed in
vacuo, and the resulting yellow solid was crystallized from
dichloromethane/ethanol. Yield: 2.6 g of brown-yellow micro-
crystals (30%). Mp: 276 °C (dec). 31P{1H} NMR (CDCl3): δ
-70.71. 1H NMR (CDCl3): δ 7.98 (d, 3J (P,H) ) 9.5 Hz, 4H),
7.20 (dd, 3J (H,H) ) 8.5 Hz, 4J (H,H) ) 1.9 Hz, 4H), 7.06 (d,
3J (H,H) ) 8.5 Hz, 4H), 6.41 (t, 3J (H,H) ) 7.5 Hz, 2H), 6.18 (d,
solution of 2.4 g of 9-chlorodibenzo[b,d]phosphole (11.0 mmol)
in 10 mL of toluene was added. The reaction mixture was
slowly warmed to room temperature overnight. Next the
diethyl ether was removed in vacuo, and the mixture was
diluted with 100 mL of dichloromethane and hydrolyzed with
30 mL of a 10% aqueous HCl solution. The water layer was
removed, and the organic layer was dried over MgSO4. The
solvents were removed in vacuo, and the resulting yellow solid
was crystallized from 2-propanol/toluene. Yield: 1.3 g of
colorless crystals (47%). Mp: 345 °C (dec). 31P{1H} NMR (THF-
d8): δ -19.78. 1H NMR (THF-d8): δ 8.1 (dd, 3J (H,H) ) 7.5
Hz, 4J (H,H) ) 1.6 Hz, 4H), 8.09 (d, 3J (H,H) ) 7.5 Hz, 4H),
7.45 (t, J (H,H) ) 7.5 Hz, 4H), 7.36 (t, J (H,H) ) 7.2 Hz, 4H),
6.95 (bs, 2H), 6.32 (bs, 2H), 1.99 (s, 6H). 13C{1H} NMR
(CDCl3): δ 153.38 (t, 11.6 Hz), 143.91 (s), 142.31 (t, 2.3 Hz),
134.53 (s), 131.36 (t, 6.2 Hz), 129.99 (s), 129.03 (s), 128.54 (s),
128.25 (s), 127.38 (t, 3.1 Hz), 126.24 (vt, 13.1 Hz), 121.38 (s),
20.41 (s). Anal. Calcd for C38H26O3P2S: C, 77.01; H, 4.42.
Found: C, 76.85; H, 4.37.
4,5-Bis(9-d iben zo[b,d ]p h osp h olyl)p h en oxa zin e (DBP -
Nixa n tp h os, 10). At 0 °C 6.8 mL of n-butyllithium (2.5 M in
hexanes, 16.9 mmol) was added to a stirred solution of 2.2 g
of 9-(tert-butyldimethylsilyl)phenoxazine (7.4 mmol) and 2.6
mL of TMEDA (16.9 mmol) in 100 mL of diethyl ether. The
resulting solution was warmed to room temperature overnight.
The reaction mixture was cooled to -78 °C, and a solution 3.7
g of 9-chlorodibenzo[b,d]phosphole (16.9 mmol) in 30 mL of
toluene was added. The reaction mixture was slowly warmed
to room temperature overnight. Next the diethyl ether was
removed in vacuo, and the mixture was diluted with 100 mL
of dichloromethane and hydrolyzed with 30 mL of a 10%
aqueous HCl solution. The water layer was removed, and the
organic layer was dried over MgSO4. The solvents were
removed in vacuo, and the resulting brown paste was dissolved
in 100 mL of THF. n-(Bu)4NF‚3H2O (4.22 g, 16.2 mmol) was
added, and the reaction mixture was stirred for 2 days at room
temperature. Then 30 mL of brine and 100 mL of dichlo-
romethane were added. The water layer was removed, and the
organic layer was dried on MgSO4. The solvents were removed
in vacuo, and the resulting yellow solid was crystallized from
dichloromethane/ethanol. Yield: 1.6 g of yellow crystals (43%).
Mp: 268 °C (dec). 31P{1H} NMR (THF-d8): δ -20.52. 1H NMR
(THF-d8): δ 8.22 (bdd, 3J (H,H) ) 7.2 Hz, 4J (H,H) ) 1.2 Hz,
4H), 7.99 (d, 3J (H,H) ) 7.5 Hz, 4H), 7.44 (t, 3J (H,H) ) 7.2 Hz,
4H), 7.37 (t, 3J (H,H) ) 7.2 Hz, 4H), 6.41 (td, 3J (H,H) ) 7.5
Hz, 3J (P,H) ) 1.2 Hz, 2H), 6.30 (d, 3J (H,H) ) 7.8 Hz, 2H), 5.93
(bdd, 3J (H,H) ) 7.5 Hz, 4J (H,H) ) 1.2 Hz, 2H), 5.5 (s, 1H).
13C{1H} NMR (THF-d8): δ 147.58 (t, 10.8 Hz), 144.98 (s),
143.11 (s), 133.92 (s), 132.45 (t, 13.1 Hz), 129.45 (s), 128.24
(s), 125.04 (s), 124.63 (vt, 12.6 Hz), 122.78 (s), 122.34 (s), 115.03
(s). Anal. Calcd for C46H35O3P3: C, 78.97; H, 4.23; N, 2.56.
Found: C, 79.11; H, 4.33; N, 2.41.
3
3
3J (H,H) ) 7.5 Hz, 2H), 5.93 (dd, J (H,H) ) 7.5 Hz, J (H,H) )
1.8 Hz, 2H), 4.76 (bs, 1H), 2.37 (s, 12H). 13C{1H} NMR
(CDCl3): δ ) 155.38 (s), 136.63 (vt, 21.9 Hz), 135.01 (s), 133.87
(t, 5.5 Hz), 133.67(t, 2.4 Hz), 132.56 (s), 131.91 (s), 125.01 (s),
124.82 (s), 118.67 (s), 118.25 (s), 114.57 (s), 20.83 (s), 20.82
(s). Anal. Calcd. for C40H31NO3P2: C, 75.58; H, 4.92; N, 2.20.
Found: C, 75.14; H, 4.89; N, 2.31.
3
4
4,5-B i s (2,8-d i m e t h y l-10-p h e n o x a p h o s p h i n o )-10-
p h en ylp h en oxa p h osp h in e (P h osxa n tp h en oxa p h os, 8). At
room temperature 23 mL of phenyllithium (1.8 M in hexanes,
41.4 mmol) was added to a stirred solution of 1.5 g of
10-phenylphenoxaphosphine (5.4 mmol) and 6.24 mL of TME-
DA (40.7 mmol) in 30 mL of diethyl ether. The dark brown
reaction mixture was heated at reflux temperature for 2 days.
Then the reaction mixture was cooled to -78 °C, and a solution
of 10.7 g of 10-chloro-2,8-dimethylphenoxaphosphine (40.7
mmol) in 80 mL of toluene was added. The reaction mixture
was slowly warmed to room temperature overnight. Next the
diethyl ether was removed in vacuo, and the mixture was
diluted with 120 mL of dichloromethane and hydrolyzed with
40 mL of a 10% aqueous HCl solution. The water layer was
removed, and the organic layer was dried over MgSO4. The
resulting solid was filtered through silica gel (eluent: toluene).
The solvent was removed in vacuo, and the resulting white
solid was purified by several washings with boiling hexanes.
Yield: 1.5 g of a white powder (37.9%). Alternatively, crystals
can be obtained from dichloromethane/ethanol. Mp: 256-258
°C. 31P{1H} NMR (CDCl3): δ -53.08, -70.38 (ratio 1:2). 1H
NMR (CD2Cl2): δ 7.93 (m, 2H), 7.84 (m, 2H), 7.40 (ddd, 4J (P,H)
Hyd r ofor m yla tion . In a typical experiment the autoclave
was charged with a 8.5 mL solution of Rh(CO)2(acac) and 5
equiv of ligand in toluene. After purging the solution three
times with CO/H2 (1:1), the reactor was brought to 16 bar of
CO/H2. Next the autoclave was heated to 80 °C. After 1 h at
80 °C the substrate and internal standard were charged to
the reaction mixture by overpressure of CO/H2 (1:1).
Deu ter iofor m yla tion . In a typical experiment the auto-
clave was charged with a 8.5 mL solution of Rh(CO)2(acac) and
5 equiv of ligand in toluene. After purging the solution three
times with D2, the reactor was brought to 8 bar of D2 and
further pressurized to 16 bar with CO. Next the autoclave was
heated to 80 °C. After 1 h at 80 °C the substrate was charged
to the reaction mixture by overpressure of CO. Samples of the
crude reaction mixture were analyzed by GC. Thereafter the
reaction mixture was distilled to separate the hexenes from
the aldehydes. Deuterium contents were determined by 2H
NMR; data identical to the data reported by Casey et al. were
obtained.32
3
4
) 10.5 Hz, J (H,H) ) 6.3 Hz, J (H,H) ) 2.4 Hz, 2H), 7.23 (m,
5H), 7.14 (d, 3J (H,H) ) 7.2 Hz, 4H), 7.09 (m, 4H), 6.96 (m,
4H), 2.34 (s, 6H), 2.32 (s, 6H).13C{1H} NMR (CD2Cl2): δ 157.05
(t, 10.3 Hz), 154.18 (d, 6.8 Hz), 140.17 (d, 21.2 Hz), 136.24 (d,
35.7 Hz), 136.02 (s), 135.74 (t, 20.61 Hz), 135.59 (t, 19.9 Hz),
133.52 (dt, 12.5 Hz, 5.6 Hz), 132.63 (s), 132.46 (s), 132.36 (s),
130.54 (vt, 17.1 Hz), 129.46 (s), 129.03 (d, 6.9 Hz), 124.94 (d,
11.7 Hz), 119.97 (d, 4.8 Hz), 118.29 (t, 2.8 Hz), 118.01 (d, 11.1
Hz), 20.95 (s), 20.90 (s). Anal. Calcd for C46H35O3P3: C, 75.82;
H, 4.84. Found: C, 75.89; H, 4.74.
4,5-Bis(9-diben zo[b,d]ph osph olyl)-2,7-dim eth ylph en ox-
a th iin (DBP -Th ixa n tp h os, 9). At 0 °C 4.4 mL of n-
butyllithium (2.5 M in hexanes, 11.0 mmol) was added to a
stirred solution of 1.1 g of 2,7-dimethylphenoxathiin (4.7 mmol)
and 1.7 mL of TMEDA (27.4 mmol) in 25 mL of diethyl ether.
The resulting solution was warmed to room temperature
overnight. The reaction mixture was cooled to -78 °C, and a

Hydroformylation of Internal Olefins
Organometallics, Vol. 22, No. 25, 2003 5369
High -P r essu r e NMR Exp er im en ts. In a typical experi-
ment a solution of 5.2 mg of Rh(CO)2(acac) (20 µmol) and 1.1
equiv of ligand in 1.5 mL of toluene-d8 was pressurized to 16
bar of CO/H2. In the NMR spectrometer the tube was heated
to 40 °C until complete formation of the rhodium hydride was
observed. NMR spectra at different temperatures were re-
corded.
40 °C. Catalyst formation was monitored in time. Hydride-
deuterium exchange was carried out by cooling the autoclave
to room temperature, venting the gases, and purging the
solution three times with 5 bar of D2. Next the autoclave was
pressurized to 5 bar D2, and finally the autoclave was brought
to 10 bar by further pressurization with CO.
High -P r essu r e F T-IR Exp er im en ts. In a typical experi-
ment the high-pressure IR autoclave was charged with a
solution of 4 mg of Rh(CO)2(acac) and 1.1 equiv of ligand in
15 mL of cyclohexane. The autoclave was purged three times
with 10 bar CO/H2 (1:1), pressurized to 16 bar, and heated to
Ack n ow led gm en t. Financial support from Celanese
Chemicals Europe, Gmbh, Germany, is gratefully
acknowledged.
OM034012F